On:

NASA selected SpaceX, Blue Origin and six other companies to compete for lunar cargo deliveries under a $10 billion contract framework, while awarding about $1.3 billion in launch and in-space transportation task orders for its Moon Base program.

The agency announced the awards Saturday at an aerospace expo at its Kennedy Space Center in Florida, saying in a release it is expanding commercial robotic deliveries, launch services and surface infrastructure for a sustained presence on the Moon’s South Pole. “We are building the Moon Base mission by mission,” NASA Administrator Jared Isaacman said.

The $10 billion cargo framework

The eight companies, Blue Origin, Firefly Aerospace, Intuitive Machines, ispace U.S., Lockheed Martin Space, Northrop Grumman Space Systems, SpaceX and Voyager Technologies, are eligible to compete for future lunar delivery missions under the new CLPS 2 contract. It is a 10-year, multiple-award framework with a combined maximum value of $10 billion; companies compete for individual delivery orders under it rather than receiving the money up front. NASA also raised the ceiling of the current CLPS contract from $2.6 billion to $4.2 billion, keeping it in place through November 2028, with deliveries running through November 2033.

Launch and transport task orders

NASA awarded six task orders worth about $1.3 billion combined to Arrow Science and Technology, Blue Origin and Impulse Space to integrate and transport Moon Base cargo. Planned missions include delivering the CAPSTONE 02 spacecraft to deep space, sending MoonFall drones to scout the lunar surface, and testing navigation and cargo transfer between Earth and Moon orbits, with launches aboard Blue Origin’s New Glenn, Relativity Space’s Terran R and SpaceX’s Falcon 9.

SpaceX and Blue Origin roles

SpaceX was selected for a Cislunar Communications Deployment Demonstration under an existing contract, launching communications satellites aboard a Starship to cislunar space in 2028 to test future lunar communications networks.

Blue Origin won a $94 million base contract, with a performance option worth an additional $140 million, to deliver NASA’s nuclear-powered PROMISE rover to the South Pole using its Mark 1 lander. The award expands a task order from May 2026.

In plain words: NASA is hiring private companies as a delivery service for the Moon instead of building everything itself. The $10 billion is a ceiling the companies compete for over 10 years, not money handed out now.

JBizNews Desk | Washington, D.C.

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

NASA plans to send a nuclear-powered rover to the Moon’s South Pole in late 2028 to scout for water ice and other resources ahead of its first Moon base.

The agency laid out the mission on its page for the rover, called PROMISE, short for Polar Rover for Observation, Mapping, and In-Situ Exploration. The SUV-size vehicle will be built by NASA’s Jet Propulsion Laboratory in Southern California for a late-2028 launch and an 18-month primary mission.

Built from Mars rover parts

PROMISE will be made largely from proven hardware that supported NASA’s Curiosity and Perseverance Mars rovers, including electronics, body structure and the driving system from a duplicate version of Curiosity used as a test bed. Power will come from a flight-spare nuclear generator, known as an MMRTG, which lets the rover work through frigid lunar nights and inside permanently shadowed craters where scientists believe ice may be preserved.

A scout for the Moon base

A commercial cargo lander will deliver PROMISE to the South Pole. Its instruments include a drill to pull up soil from about three feet down to study water ice, spectrometers to analyze its chemistry, temperature and radiation sensors, and cameras. NASA calls the mission part of the first phase of its Moon Base program, a series of robotic missions to scout, experiment and prepare for astronauts to live and work on the surface.

In plain words: This rover is due to launch in late 2028 to scout for water ice near the Moon’s South Pole before astronauts arrive.

JBizNews Desk | Pasadena, California

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

File photo of a Tesla Cybertruck on a mountain road, not the Texas factory.

On:

Photo: Stephen Leonardi / Pexels

Tesla will add an extra Cybertruck production shift to try to clear outstanding orders before it dismantles the pickup’s Texas line around the middle of next year to make room for the Cybercab, Elon Musk said Saturday.

“The Cybertruck order backlog is currently far more than Tesla can make before dismantling the line for Cybercab expansion around June/July next year,” Musk wrote in a post on X. “We will add an extra shift to try to get all orders done in time.”

Tesla reported 8,295 deliveries in its Other Models category in the third quarter, down from 12,364 in the second quarter; the category is not a standalone Cybertruck total.

Musk said Friday on X that Tesla plans to wind down Cybertruck production at its Texas Gigafactory around mid-2027 to make room for more Cybercab production. He called the truck the “best vehicle Tesla has ever made” and the move the “end of an era.”

The production shift would redirect space at the Austin plant from Tesla’s stainless-steel electric pickup toward the vehicle at the center of its robotaxi ambitions.

The future of Cybertruck manufacturing beyond that factory remains unclear: production could move elsewhere or end. The Texas plant is the only known Tesla facility equipped to manufacture the model’s specialized components.

Cybercab takes more of the factory floor

Musk’s plan puts the factory’s next phase behind Cybercab, the autonomous vehicle he has made central to Tesla’s push into self-driving transport. The timetable is around the middle of 2027, rather than an immediate halt to the truck line.

Cybertruck lead engineer Wes Morrill pointed to a technical link between the two vehicles. “Cybertruck was an incredible technology development platform and lives on through Cybercab,” he said on X.

The pickup launched in 2023 as Tesla’s entry into a truck market long dominated by established manufacturers such as Ford. It has struggled to reach the mass-market audience Musk once envisioned, while facing recalls, quality-control issues and repeated price changes.

The planned Texas wind-down is another step in Musk’s effort to turn Tesla from a vehicle maker into a company built around robotics and self-driving services, with Cybercab intended as the main autonomous model in a wider driverless network.

JBizNews Desk | Austin

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Binance, the world’s biggest cryptocurrency exchange, said Saturday that Chief Financial Officer Kaiser Ng is leaving after more than five years and that Justin Yek will become its new finance chief.

Co-Chief Executive Yi He said in the statement that Ng “has decided to pursue a new venture after more than five years at Binance, with a transition plan implemented over the last half year.”

A handover planned for months

Yi He said the change was planned over the last six months. Ng leaves after more than five years with the company. Yek steps into the role as Binance works through the fallout of the largest theft in crypto history.

Days after a record theft

The announcement comes about a week after hackers stole $6.3 billion from the exchange. Binance froze withdrawals for several days, and users pulled out a net $3.6 billion in the first hours after accounts reopened, a report said. The company and the FBI blamed a North Korean group called TraderTraitor.

Binance has said it will cover the stolen funds itself. “Binance will bear the financial impact of the stolen funds,” the company said. “These losses will not be passed on to users.”

In plain words: Binance’s money chief is leaving and the transition plan was put in place over the last six months. The company says customers will not pay for last week’s giant hack.

JBizNews Desk | Dubai, United Arab Emirates

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

SpaceX says its next Starship test flight will try to do something no rocket has done before: fly all the way around Earth, then bring the ship back to its Texas launch site for a catch by the launch tower’s mechanical arms.

The company laid out the Flight 15 plan in an update Saturday. Bringing the upper stage home is the last big step toward reusing the entire rocket, which SpaceX says would let it fly again quickly, carry more satellites to orbit and lower costs for Starlink and launch customers. On earlier flights, Starship ended its missions with splashdowns in the ocean.

A 21-hour trip around Earth

Starship will first launch on a path that falls safely into the Indian Ocean if anything goes wrong early. Once flight controllers confirm the vehicle is healthy, they will give a “go” for orbit, and a single engine will push it to orbital speed.

In orbit, Starship will deploy a batch of Starlink V3 satellites. Three of them are modified to take pictures of Starship’s heatshield, one of several checks on whether the ship is ready to come home. The ship will then coast in orbit for about 21 hours while controllers run constant health checks. If a problem appears, they can end the mission early with splashdowns in pre-cleared parts of the Pacific Ocean near Hawaii or off Chile.

Coming home over land

To reach Starbase, Starship must fly over inhabited areas of northern Mexico and southern Texas, something its earlier flights avoided. SpaceX says it will send the ship back only if it meets return conditions approved by federal aviation regulators. The ship will steer with its four flaps along a path over sparsely populated areas, weaving around the major cities of the Rio Grande Valley before a final turn toward the launch site. People in Cameron County may hear one or more sonic booms as it arrives.

The catch

Over Starbase, Starship will flip vertical, fire its engines and slow to a near hover between the tower’s arms at Pad 2. The arms will then close rapidly to catch the ship by two fittings below its forward flaps. If anything is wrong with the ship or the tower, controllers can break the vehicle up high in the atmosphere or steer it into the Gulf instead.

SpaceX calls full recovery of an orbital rocket the “Holy Grail” of rocketry because it would allow a vehicle to fly again rapidly, like an airplane.

In plain words: SpaceX wants to catch its biggest rocket with giant arms instead of dumping it in the ocean. If it works, rockets could fly again and again, making satellite internet and space launches cheaper.

JBizNews Desk | Starbase, Texas

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

By: JBizNews Desk 🌎

Nvidia’s China chip business faces fresh scrutiny after a New York Times investigation reported that a Chinese company approved to assemble AI servers using its H200 processors had an empty listed office and owners linked to a US-blacklisted fund.

The New York Times reports that the Commerce Department cleared New Matrix earlier this year to assemble the advanced chips into computers used for artificial intelligence. When reporters visited its listed address in a Beijing industrial park, they found an empty office with exposed wiring. The building directory did not list the company, and the property’s managers said they had no record of it.

Chinese corporate records linked New Matrix’s owners to Wise Road, a private equity firm placed on the Commerce Department’s trade blacklist in 2024. The department said Wise Road had helped the Chinese government acquire companies with strategically sensitive semiconductor technology. Two analysts who reviewed the records told the Times they believed New Matrix could be a front for firms connected to the fund.

The investigation examines the screening behind US approvals for Chinese companies to buy or handle Nvidia’s H200 chips. President Donald Trump allowed sales of the processors to China in December, with chips routed through the United States so Washington could collect a 25% tariff on their value.

Roughly two dozen Chinese companies have received approval to buy or handle H200 chips, according to people familiar with the approvals cited by the Times. The group includes large companies such as Baidu, BYD and Tesla Shanghai, alongside firms that assemble chips into servers for other customers.

Commerce and Nvidia defend their safeguards

A spokesperson for the Commerce Department’s Bureau of Industry and Security told the Times that the agency carefully reviews applications for national-security risks. The bureau said it imposed $324 million in penalties for export violations in 2025 and hired more than 90 enforcement officers this year.

Nvidia spokesperson John Rizzo said “it makes no sense for the Chinese military to depend on American technology.” He said the company sells its first, best and most technology in the United States, and that its limited H200 exports support American jobs and manufacturing.

Rizzo also said Beijing has restricted H200 imports to favor China’s own chipmakers. H200 shipments to customers based in China accounted for less than 1% of Nvidia’s data-center revenue in the quarter ending in July, according to financial disclosures cited by the Times. In August, Nvidia said its forecast for the current quarter assumed no data-center sales to China.

JBizNews Desk | Washington, DC

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

By: JBizNews Desk 🌎

SoftBank and ABB are exploring a way to separate ABB’s Chinese robotics operations from a $5.4 billion takeover, a move that could keep the broader sale alive after Chinese authorities blocked the acquisition in China, Bloomberg reported Saturday.

The proposal would put the China business into a separate entity with different ownership, according to people familiar with the matter cited by Bloomberg. China accounts for about 30% of ABB’s robotics business and is its largest robotics market.

ABB confirmed Thursday that SoftBank had informed it that China’s Office of the Working Mechanism for Foreign Investment Security Review prohibited the acquisition of ABB Robotics’ business in China. The companies agreed in October 2025 on an enterprise value of $5.375 billion for the robotics division.

The Swiss engineering group said the transaction remained subject to regulatory approvals and customary closing conditions, with SoftBank managing the approval process. ABB said the companies were in close dialogue and working constructively on the transaction.

A China carve-out would change the structure of a major robotics transaction at a time when governments are scrutinizing foreign ownership of advanced technology.

JBizNews Desk | Zurich, Switzerland

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Microsoft says a financially motivated hacking group based in Cambodia has been using an AI-powered security-testing framework to attack financial organizations, adding to evidence that tools built for legitimate testing are being used for cybercrime.

For banks and other financial businesses, the warning is about the speed of attacks against exposed systems and stolen accounts. Microsoft says the tools can move through parts of an intrusion quickly, making early detection and protection of internet-facing systems important.

In its Friday research report, Microsoft said the group it tracks as Storm-3358 began using the ARTEX framework against financial-sector organizations in October. Its targets were mainly in Southeast Asia, with a smaller number of large financial organizations elsewhere.

Microsoft said the group combined ARTEX with conventional automated tools to scan internet-connected systems, exploit security flaws and install remote-access malware.

The company also said ARTEX-related attacks more broadly can involve stolen credentials and compromised cloud identities. Its report includes detection guidance and indicators for security teams to investigate.

JBizNews Desk | Redmond, Wash.

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

AI illustration of two server-rack rows in a dark data center.

On:

Google Cloud has opened its A5X computing systems, built on Nvidia’s new Vera Rubin chips, to customers. The systems became generally available Friday, Nov. 6, so companies can now rent them to train and run AI models.

Google Cloud says in its announcement that A5X instances are bare-metal machines, meaning customers get the physical servers rather than a shared slice. They run on Nvidia’s Vera Rubin NVL72 platform, a rack-sized system that works as one.

Built for AI agents

Google says the systems are aimed at “agentic” AI, software that carries out long, multi-step jobs on its own. Companies running those jobs need to move huge amounts of data quickly between chips and storage, and Google says A5X is designed around that. It pairs the machines with its Kubernetes tools for scheduling large AI jobs across many servers.

A customer’s view

Stephan Pretorius, chief technology officer of advertising group WPP, said the systems let its creative and production teams “direct intelligent agents that perform frame-accurate quality checks and adapt brand assets seamlessly.”

In plain words: Nvidia’s newest AI machines are now for rent on Google’s cloud. Companies that want to build or run big AI programs no longer have to buy their own.

JBizNews Desk | Mountain View, California

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

AI illustration of a man using a smartphone beside a mug and tablet in a kitchen.

On:

Meta is rolling out upgraded AI models to its Muse and Meta AI assistants that are built to remember what matters from past chats and keep going when a task hits a snag.

The company announced the update Friday. The new models were trained using Muse Wonder, the company’s most capable model series, which it introduced a few weeks ago. They add two features called Active Memory and Active Focus.

Remembering what matters

Active Memory is meant to help the assistant recall earlier information when it becomes useful. Meta says it trained the models to tell temporary information from permanent, and what is still current from what has gone stale. In Meta’s own tests, the new Muse model used critical context correctly in 90% of tasks, up from 76% for the previous model. Meta also says it cut the forgetting of critical information by 2.4 times.

Sticking with the task

Active Focus is meant to keep Muse working toward a goal when something unexpected gets in the way. Meta says the models try harder, recover, or find another route. In its own tests, Muse completed everyday tasks with unexpected problems 67% better than before. The models can also juggle several jobs, decide when to work alone or with you, and take a correction in the middle of a task without getting confused.

Meta AI gets the same upgrade

In Meta AI, the company says the same changes make it a better partner for thinking things through. It can break a big problem into steps, ask sharper follow-up questions, and keep up when you change direction mid-conversation. On Meta’s Personal Superintelligence Index, the new Muse model scored 47%, up from 38%.

In plain words: Meta says its assistants should now remember the useful parts of your past chats and not give up when a task gets tricky. The scores come from Meta’s own tests.

JBizNews Desk | Menlo Park, California

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

AI illustration of a generic electric car connected to a home charger beside a battery in a garage.

On:

Tesla Energy says its Smart Charging feature can now coordinate home charging for electric vehicles from other brands when used with a Powerwall battery and Wall Connector charger.

For households that own Tesla’s home-energy equipment but drive another brand’s EV, the feature offers a way to time charging around solar generation and electricity rates. It requires the Powerwall-and-Wall-Connector setup Tesla named, rather than the Tesla app alone.

In a November 6 announcement, Tesla Energy said Smart Charging considers a home’s solar production, energy goals and driving needs, then automatically chooses charging times. The company described it as “now compatible with any EV” and part of Tesla Home in the app.

Tesla described two charging options. Savings Mode enables Low-Cost Charging, which takes account of solar production and utility rates. Self-Powered Mode enables Charge on Solar, which sends surplus solar energy to the vehicle instead of exporting it to the grid.

The feature is part of Tesla’s wider home-energy controls. On its Tesla Home page, the company says its software plans energy use around household demand, solar output and local electricity prices.

JBizNews Desk | Austin, Texas

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

AI illustration of a wafer-measuring tool in a clean chipmaking laboratory.

On:

Nova, an Israeli maker of chip-measuring equipment, reported record sales and profit for the third quarter and forecast higher sales for the final three months of the year. Its shares rebounded in Friday midday trading, up about 10.9%.

Nova said Thursday in its results that revenue was $287.8 million, up 28% from a year earlier. Profit under standard accounting rules was $90.0 million, or $2.64 a share. Excluding certain items such as stock pay and acquisition costs, profit was $98.4 million, or $2.86 a share. Both profit figures were records, the company said.

Memory and advanced packaging drive sales

Nova sells tools that measure and inspect computer chips while they are made, so factories can catch problems early. The company said it had record sales from memory chips and from advanced packaging, a way of stacking and connecting chips that is used in high-speed AI memory. It said memory demand came from its tools that measure the size and materials of chip layers, and that packaging sales grew as customers added capacity for high-bandwidth memory and advanced logic chips.

What Nova expects next

For the fourth quarter, Nova expects revenue of $305 million to $315 million and adjusted earnings of $3.03 to $3.18 a share. Chief Executive Gaby Waisman said the company is confident it can win business from the biggest changes in chip technology because it can better see its customers’ investment plans.

In plain words: Nova sells the tools chipmakers use to check their work. Chipmakers are spending more on memory and advanced chips, so Nova sold more, earned more, and expects to sell more next quarter.

JBizNews Desk | Rehovot, Israel

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Illustration of a quantum-computing chip on a round base.

On:

Physical Superintelligence, known as PSI, says an AI research system it developed found a proof of the Aaronson-Ambainis conjecture, a longstanding mathematical problem about when quantum computers can outperform ordinary computers.

For researchers developing AI tools for mathematics, the claim offers a test of whether an AI system can produce new results that people and software can check. For quantum-computing researchers, the result concerns the limits of a particular kind of speed advantage, rather than a new computer or a demonstration of faster hardware.

In its November 6 research announcement, PSI said its system ran on the company’s infrastructure using instances of a non-public Anthropic model. It said the system had a complete proof during the third day of the run, then improved the result twice. According to PSI, the Lean proof assistant accepted the final version.

The conjecture concerns how many times a computer must look at parts of its input to produce an answer. PSI says its proof establishes that, for most randomly chosen inputs in the setting studied, an ordinary computer can approximate a quantum computer’s answer probabilities without needing exponentially more such looks.

That is a statement about access to input data, not a general comparison of total running time. The original Aaronson-Ambainis paper describes a model that counts queries while ignoring other computational steps.

PSI has published a paper, formal proofs and verification materials. Its announcement credits David Turturean with developing and running the system, reviewing the paper’s draft and taking responsibility for its correctness. PSI says Jim Halverson audited the proof and Phiala Shanahan commented on its Lean formalization.

JBizNews Desk | Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Illustration of a rocket and launch tower viewed across coastal wetlands.

On:

Elon Musk said government revenue would account for less than 10% of SpaceX’s revenue this quarter and less than 5% a year from now, forecasting a smaller share of the company’s business from government customers.

The forecast came in a Friday post in which Musk also offered to cancel government contracts that officials did not want. His comments were an offer, not an announcement that any contract had been canceled.

For SpaceX investors, the figures describe Musk’s view of the company’s future customer mix. They do not establish a drop in government revenue in dollar terms, since a smaller share can also result from other parts of a business growing faster.

In the November 6 post, Musk wrote that “government revenue will be less than 10% of SpaceX revenue this quarter and less then 5% a year from now.”

He added: “Happy to cancel any that the government doesn’t want. Just say the word!”

Musk was responding to a post proposing that he offer to cancel federal contracts criticized by lawmakers. He described SpaceX’s government contracts as offering both the best product and the lowest price, a characterization made by Musk rather than an independent comparison of competing bids.

The post did not identify contracts to be ended, list a dollar value for possible cancellations or announce that a government agency had accepted the offer. The revenue percentages are Musk’s forecast, not reported quarterly results.

JBizNews Desk | Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Illustration of a monitor, microscope and sample tube rack on a laboratory bench.

On:

OpenAI introduced a new test of how artificial intelligence handles everyday laboratory evidence, reporting stronger results from its GPT-6 Astra model while warning that higher scores do not yet mean reliable judgments in the lab.

The benchmark, developed with enzyme-design company Ligo, is called BioBasicsBench. It tests AI agents on 100 tasks drawn from routine laboratory work, including interpreting experimental images, spotting errors and recognizing when the evidence cannot support a conclusion.

For researchers and biotechnology businesses, the issue is practical: a mistaken reading of an experiment or a missed problem with its controls can send scientists toward the wrong conclusion and waste time on follow-up work.

In its November 6 announcement, OpenAI said GPT-6 Astra earned an average of 66.7% of the available scoring credit at the test’s “xhigh” reasoning setting, compared with 43.8% for GPT-5.6 Sol. Those figures are average benchmark scores, not the percentage of experiments completed without mistakes.

Testing raw evidence, not polished results

The tasks use material such as gel images, microscope images, sequencing files, instrument readouts and videos of laboratory work. Scientists built the questions from their own experimental data and paired them with expert-written grading criteria.

OpenAI said the evaluation covered seven models and 37 combinations of models and reasoning settings. Each combination made five attempts on each of the 100 tasks, producing 18,500 results.

The company also documented continuing mistakes. In one example, Astra selected the wrong colonies from experimental results after an assumption overrode the comparison supported by the controls. In another, it overcounted colonies by approximately 12% relative to the benchmark reference.

OpenAI said the benchmark is intended to track progress in interpreting raw data and checking whether conclusions are supported. The results do not establish that an AI system can replace a scientist’s review.

JBizNews Desk | Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Illustration of three server cabinets in a data center

On:

Businesses and users who connect OpenAI tools to other services are the focus of a new security partnership. OpenAI said Friday it will work with Hacktron to find weaknesses where its systems meet outside services.

OpenAI engineers and Hacktron security researchers will jointly choose high-priority areas to test, focusing on interactions with third-party services, the company said in its partnership announcement. OpenAI is also expanding Hacktron’s access to its frontier models through Daybreak Red.

The work targets a growing challenge for AI systems that connect apps and services: a weakness can arise in the connections between systems, rather than within one app alone. OpenAI said it will use the testing to strengthen its defenses.

Hacktron builds AI agents for code review, penetration testing and red teaming. OpenAI said the researchers’ work has uncovered vulnerabilities in the way applications, cloud infrastructure and identity systems interact.

OpenAI said it first became familiar with Hacktron through its Security Bug Bounty program and was impressed by vulnerabilities the researchers had found and disclosed.

“The future of cybersecurity won’t be defined by who has the largest security team, but by who can best combine human creativity with AI-driven capabilities,” Hacktron CEO Zayne Zhang said in the announcement.

The partnership adds to OpenAI’s security work announced earlier this week, including plans to work with Cathedral researchers in its next security sprint and higher rewards for independent bug hunters.

JBizNews Desk | Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Illustration of a communications satellite with laser links above Earth

On:

Elon Musk says SpaceX could fit about 100 laser links on each future Starlink satellite, outlining an engineering goal he believes could help the orbital network compete with undersea internet cables.

The goal concerns links that carry data between satellites, which is separate from the internet speed delivered to a customer’s home.

In an evening post Friday, Musk wrote: “If we shrink laser size, assuming 10X closer proximity, and have small gimbal angle for in-plane links, maybe we can get to 100 lasers/sat.”

The post names smaller laser equipment, a narrower aiming range and closer spacing between satellites as the conditions for reaching that goal.

In an earlier post, Musk described a network of more than 100,000 satellites, each carrying roughly 100 lasers operating above 10 terabits per second. He said that configuration “should be able to outcompete undersea cables.”

Musk also said light travels about 40% faster through space than through cable and that a space-based route can be more direct. Those are Musk’s comparisons for his proposed network.

JBizNews Desk | Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Archive photo of a Falcon9launchfromVandenbergonOctober15,2025.Nottonight'sUSSF-432launch.IanHawkes/U.S.SpaceForce.

On:

Archive: Ian Hawkes / U.S. Space Force

SpaceX launched the USSF-432 mission from California Friday evening and confirmed its Falcon 9 first stage landed on the droneship Of Course I Still Love You in the Pacific Ocean. SpaceX announced liftoff at 7:49 p.m. Eastern and the landing at 7:59 p.m.

The rocket took off from Vandenberg Space Force Base. SpaceX posted the liftoff and the landing on X.

A booster flying for the eighth time

Before the launch, SpaceX said this was the eighth flight for the first stage, which earlier flew a GPS III-8 mission and six Starlink missions, according to its mission page. A first stage is the rocket’s bottom section. Landing it on a ship lets SpaceX use it again.

Sonic booms were possible in California

SpaceX had said residents of Santa Barbara, San Luis Obispo and Ventura counties might hear one or more sonic booms during the launch, depending on weather and other conditions.

In plain words: SpaceX sent a rocket up from California on Friday night, and the lower part of the rocket came back down and landed on a ship in the ocean so it can fly again.

JBizNews Desk | Vandenberg Space Force Base, California

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Archive: Falcon 9 launches Bluebird 1-5 satellites from Cape Canaveral Space Force Station, Florida, September 12, 2024. DeAnna Murano / U.S. Space Force.

On:

Archive: DeAnna Murano / U.S. Space Force

Elon Musk says SpaceX’s yearly revenue pace could top $500 billion by December 2027, up from a pace of more than $100 billion he expects this December. He called the 2027 figure his best guess, not a promise.

Musk made the remarks Friday in a conversation at the Baron Investment Conference. He said SpaceX is tracking to an annual rate of revenue “in excess of 100 billion” dollars in December. He then said “it seems quite likely that our annual recurring revenue in December of next year will exceed 500 billion,” and added: “This is not a promise… This is my best guess.”

What the number means

Both numbers are forecasts from Musk, not revenue SpaceX has reported. Musk described a revenue rate at a particular point in time, rather than total sales over a full year. Reaching that rate in December 2027 would not mean SpaceX recorded $500 billion in revenue during 2027.

Who is affected

SpaceX shareholders, and anyone weighing whether to buy shares, are the people most affected, because they are judging forecasts like this against what the company has actually reported.

In plain words: Musk is predicting that SpaceX will be selling at a pace of more than $500 billion a year by the end of next year. He says it is his best guess, not something he is promising.

JBizNews Desk | New York

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Archive photo of a Rumo freight train in Pinhais, Paraná, Brazil, March 2026.

On:

Archive: Matt9019 (CC0)

Brazilian rail operator Rumo is using Starlink and 4G connections across more than 500 locomotives and railcars to keep trains in touch with its control center where cellular coverage drops, according to connectivity supplier Autotrac.

Autotrac’s commercial director, Márcio Toscano, described the setup Thursday at the Movecomm industry event in Brazil, according to an event report. The count is of pieces of equipment, locomotives and railcars, not of whole trains.

How the hybrid connection works

Autotrac calls the service hybrid connectivity because it combines satellite and cellular networks. It was developed with Starlink and uses 4G networks from carriers including TIM, Claro and Vivo, switching between ground and satellite links to keep data moving, Toscano said. The point is real-time contact between each train and Rumo’s operations control center along routes where phone coverage is patchy.

Other uses Autotrac described

Toscano said Rumo also uses cameras with artificial intelligence to spot signs of driver fatigue. Autotrac also serves road-concession and utility vehicle fleets, including Copel in Paraná state, where vehicles act as mobile hotspots for crews. One of its next projects is high-speed Wi-Fi for passengers on intercity and charter buses.

In plain words: trains cross long stretches where phone signals fade. Rumo’s equipment can switch to a satellite link so controllers stay in touch with crews, according to the company that supplies the service.

JBizNews Desk | Brazil

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Rendering of Core Scientific's Auburn, Alabama data center, supplied by the company

On:

Rendering: Core Scientific

Core Scientific plans to leave Bitcoin mining by year-end and focus on data-center space for AI and high-performance computing, shifting a former major miner’s business toward customers such as CoreWeave and AMD.

Chief Financial Officer Jim Nygaard said on the company’s Thursday earnings call that it expects to finish winding down mining in the fourth quarter and begin 2027 as a provider of high-density colocation, meaning it rents out power and data-center space to customers. The exit is a target, not a finished move. The call transcript shows the company had nearly 50% fewer miners online at the end of September than at the end of 2025.

Colocation now makes up most of the revenue

The company’s results release shows third-quarter revenue of $232.9 million, up from $164.2 million in the second quarter and $81.1 million a year earlier. Colocation brought in $215.4 million of that. Mining the company does for itself brought in $12.3 million, and mining it hosts for others brought in $5.2 million.

The company said it was billing for 485 megawatts of capacity at the end of September. Core Scientific said it is converting its remaining mining sites to serve colocation customers.

The big profit number comes from an accounting gain

Net income was $1.178 billion, but the company said it was mainly driven by a noncash gain from a change in the value of its warrants, which move with its stock price. It is not cash earned from running the business. Adjusted EBITDA, a measure that leaves out some costs and accounting effects, was $133.6 million, up from $41.1 million in the second quarter.

What could slow the plan

Management said further growth depends on delivering more capacity for CoreWeave and AMD and on raising money to build it. Nygaard said the company is working on project bonds and other debt. The company’s stock fell 10.82% to $11.38 in Thursday’s regular session, a day when the broader market also fell, so the drop cannot be tied to the mining plan alone.

In plain words: Core Scientific is planning to stop mining Bitcoin and rent out data-center space instead. Most of its revenue already comes from that, but its plans rely on finishing new buildings and raising money to pay for them.

JBizNews Desk | Miami

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Archive photo: Jensen Huang speaks at the White House, April 30, 2025.

On:

Archive: Joyce N. Boghosian/White House (public domain).

A scheduled meeting of President Trump’s outside science and technology advisers was called off Friday, delaying a forum that includes leaders of Nvidia, Meta, AMD and Oracle as Washington debates AI development. An Energy Department spokesperson confirmed the cancellation, according to a report.

The meeting was to bring together the President’s Council of Advisors on Science and Technology, known as PCAST, an outside panel that advises the president. Members include Nvidia CEO Jensen Huang, Meta CEO Mark Zuckerberg, AMD CEO Lisa Su, Oracle Chief Technology Officer Larry Ellison and Andreessen Horowitz co-founder Marc Andreessen. Its co-chairs are venture capitalist David Sacks and White House tech adviser Michael Kratsios.

A new date is unsettled

The Energy Department’s meeting website briefly listed Nov. 13 as the new date, then the department emailed to retract it, the report said. A federal notice published Friday still lists a rescheduled meeting on Friday, Nov. 13, from noon to 1 p.m. ET. The two do not match, so the date is not settled.

What the panel was set to do

The original meeting announcement said the panel would advise the administration’s new AI task force, called the “Super Intelligence Task Force,” which includes cabinet officials, Trump appointees and White House staff, according to the report. The Federal Register notice describes the meeting as virtual and livestreamed.

In plain words: the White House’s tech advisers did not meet Friday as planned, and the next date is unclear.

JBizNews Desk | Washington

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Archive: New York Stock Exchange, April 2012. Jean-Christophe Benoist / CC BY 3.0, cropped.

On:

Photo: J-C. Benoist / Wikimedia (archive; CC BY 3.0)

Meta and four other companies have joined the list of issuer objections to OKXICE’s planned tokenized-stock market, narrowing the potential selection for investors as the venture works toward launch. The venue is a planned market for digital tokens that stand in for U.S. shares.

The venture’s public notice, dated Thursday, Nov. 5, now lists 67 issuers that have sent a notice of objection. The five added in the latest update are Meta, Palantir, S&P Global, KLA and Super Micro Computer. Apple was added earlier this week. The objections concern this proposed venue only. They are not a ban on tokenized shares, and not an SEC prohibition on any company’s stock.

What the objections mean

Companies have 30 calendar days after receiving notice to object to their shares being tokenized for trading by an unaffiliated outside party, according to a report. Venues can take up to five business days to publish objections, so the public list can trail what companies have decided. A company missing from the list has not signed on; the notice lists objections only.

A proposed market

OKXICE is a joint venture owned 50% by Intercontinental Exchange, the parent of the New York Stock Exchange, and 50% by OKC USA, the notice says. It plans to let approved retail and institutional investors buy and sell tokenized U.S. stocks against stablecoins through automated market-maker pools, which are software-run pools of assets rather than an order book. Access is limited to people who pass identity checks.

The venue would operate under the SEC’s temporary exemption for tokenized-stock trading venues and is not registered with the SEC, the notice says. Tokenization uses digital tokens to represent financial assets, and the proposed market is separate from conventional stock-exchange trading.

OKXICE co-chairman Andrew Cuomo, the former New York governor, has said companies outside the initial selection approached the venture about joining, but he did not name them or give a number, according to the same report.

In plain words: 67 companies have now objected to their shares being used on this planned tokenized-stock market, including Meta and Apple.

JBizNews Desk | New York

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Amsterdam skyline with construction cranes (illustrative, not the protest site)

On:

Archive: Haberdoedas Photography/Pexels

Dutch police said they arrested between 150 and 200 people after protesters entered the Amsterdam construction site of a planned Microsoft data center on Friday. The arrests were for trespassing. It matters to people near proposed data centers, who share the local power and water, and to companies planning AI capacity.

What happened

  • Several hundred activists occupied the site in the Sloterdijk district, halting work for the day, sprayed anti-Microsoft messages on walls under construction and shouted political slogans. Others left the site on their own.
  • Police said they first ordered protesters to hand over tools, then allowed them to a designated protest area. Protesters then forced their way onto the site, some climbing a crane, and clashed with police.
  • Police said some of those detained were suspected of damaging property, and material suspected of being stolen from the site was found on some. All those arrested were to be questioned.
  • Protest group Geef Tegengas, which means “Push Back,” says it wants construction stopped.

Who owns the site and what they say

The site owner is Pure Data Centres, which plans to invest more than $1 billion in the project. It asked the people on site to leave safely and talk with it about their concerns.

Microsoft said it understands that people feel strongly about the project and respects their right to express that peacefully, but does not condone unlawful actions or actions that put people’s safety at risk.

A wider fight

The Amsterdam protest is the start of two days of actions planned by the “Stop Dirty Data Centres” coalition, with groups in 15 or more European countries expected to join. Other planned actions include a march in Dublin, a protest outside parliament in Warsaw and a demonstration at another Microsoft data center in Denmark. The coalition says it does not want existing data centers closed, but wants major new construction paused until governments set clearer rules.

Campaigners say data centers strain electricity and water supplies. Technology companies say they are essential for AI and cloud computing.

In plain words: police arrested up to 200 people at the Amsterdam site, and the protest is the first of two days of data center protests across Europe.

JBizNews Desk | Amsterdam

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Photo: Steve Jurvetson / Wikimedia Commons,
CC BY 2.0

OpenAI has closed on about $11 billion of new funding at a $1.25 trillion valuation, people familiar with the matter said Friday, according to a report.

The people said the money was wired to the company as of Friday and is the first part of a larger round. The rest of the round could bring the total to as much as $20 billion in the coming weeks, one person said.

A smaller round than first planned

OpenAI first set out to raise $30 billion or more at a $1.4 trillion valuation, the report said, before scaling back its plans as AI stocks sold off. The company has been in talks for more than a month, after delaying its plans for an initial public offering, according to the report. An earlier report on Thursday evening said OpenAI had raised less than half of its first $30 billion target, with $11 billion received so far at the $1.25 trillion valuation.

The report said the mood around AI has shifted since the round began, because of concerns about spending on the physical infrastructure for artificial intelligence and possible new obstacles to data center construction after this week’s midterm elections.

Where the valuation stands

The report said OpenAI raised $122 billion in a March funding round at an $852 billion valuation. The reported $1.25 trillion valuation is higher than that March figure and below the $1.4 trillion first discussed for this round.

In plain words: OpenAI has reportedly collected about $11 billion from investors so far, and may get more, but less than it first hoped for.

JBizNews Desk | San Francisco

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Elon Musk shakes hands outside the Pentagon, March 2025 archive.

On:

Photo: Madelyn Keech / DOD (2025 archive; cropped)

SpaceX is aiming to begin launching its Starmind space computers in the third quarter of 2027, Elon Musk said Friday, calling the timing a guess.

Musk spoke by video link at an investment conference in New York, in an hour-long conversation with Baron Capital CEO Ron Baron, according to a report on the event. The plan matters to anyone following SpaceX and Tesla.

What Musk said about Starmind

“We expect to start launching the first Starmind computers, which we’ve co-designed with NVIDIA, probably – I’m just guessing here – but in the 3rd quarter of next year,” Musk said. “And we’ll start off slow, but then it will ramp to significant volume probably in ’28, and then very, very large volume in ’29.”

Starmind is SpaceX’s project to capture solar energy in space and deliver computing power for artificial intelligence from a satellite constellation. Musk said going into space is needed to reach very large scales of computing power because “there just isn’t enough power on the ground.”

He said two test satellites are already up. “There’s a satellite with an NVIDIA GPU up there that we launched, and there’s a satellite we recently launched with a Google TPU,” he said. “And these are operating to the best of my knowledge – they’re working fine in space.” Google’s chips are part of a research effort called Project Suncatcher.

Opening the conversation, Musk quipped: “The pitchfork level for data centers is pretty high.”

Where the idea came from

Earlier this week, Elon Musk responded to the Democrats’ midterm election sweep with a pair of sarcastic posts on X early Wednesday, suggesting that putting data centers in space was suddenly looking attractive, a report said.

“Data centers in space looking pretty appealing right now,” Musk wrote. About a minute later, just before 1:15 a.m. Eastern, he added: “Guess my election-interfering space lasers didn’t work this time 🤷‍♂️.”

The posts nodded at a real political problem for the world’s richest man: the candidates he spent record sums supporting lost badly, and the winners in his adopted home state of Texas have promised to clamp down on the data-center industry his companies are betting on.

A bad night for Musk’s candidates

Democrats won control of both the House and the Senate on Tuesday. In Texas, Democrat Gina Hinojosa defeated three-term Republican Governor Greg Abbott, and Democrat James Talarico beat Republican Ken Paxton for a U.S. Senate seat. Musk was the biggest known donor of the 2026 midterms, but the totals depend on what you count. An analysis of campaign finance records put his overall election spending at at least $474 million, with $224 million – nearly half – going to Texas alone, where his favored candidates lost. A separate tally put roughly $250 million into congressional races specifically. His America PAC spent nearly $87 million across 44 federal races, and 93 percent of that money went to races Republicans lost, the analysis said. Talarico had accused him during the campaign of trying to buy “his own personal U.S. Senator.”

Texas turned data centers into an election issue

Hinojosa campaigned on reining in the state’s data-center boom, calling for a moratorium on new facilities until lawmakers pass stronger protections for electricity supplies, water resources and local communities. “We need a moratorium on data centers until the legislature can meet to put in place real protections,” she said in an August campaign statement. Talarico has also said data centers are driving up utility bills across Texas and has promised more regulation.

For families in Texas, the fight is about monthly utility bills and the power and water their communities rely on, and the state’s incoming leaders have promised tighter rules.

What happens next is limited by politics. Republicans kept control of the Texas Legislature, so Hinojosa cannot change state law on her own. But her administration will shape the state decisions Musk’s Texas operations depend on – electricity, water, environmental permits, infrastructure and tax incentives, according to that analysis. Hinojosa has called for an end to industry tax breaks and for data-center developers to cover their full electricity costs, and her team could slow expansion with a tougher line on permits and incentives.

Abbott had already halted new data-center permits in September, pending an audit of the state’s electricity grid. The North American Electric Reliability Corporation has warned that peak summer power demand could rise by 224 gigawatts over the next decade, largely because of AI data centers, and a report says Texas is reviewing proposed data-center connections representing more than 400 gigawatts of potential demand, though many projects remain speculative.

The stakes for Musk are direct. Tesla and SpaceX are both headquartered in Texas, SpaceX operates its Starship launch site in South Texas, and Musk has announced plans for Terafab, a semiconductor project involving the two companies.

The space idea is not entirely a joke

Musk has talked for years about moving computing off the planet. In January, SpaceX filed an application with the Federal Communications Commission to launch and operate an orbital data-center system of up to one million satellites, describing it as “the first step in becoming a Kardashev II level civilization” that could harness the Sun’s energy. The FCC accepted the application for filing in February and is seeking public comment.

Musk has predicted that monthly additions to orbital computing capacity could surpass additions on Earth by late 2028, citing electricity and regulatory constraints, and argued this week that cooling AI equipment is easier in space because radiators can release heat directly. A report in June said SpaceX aims to test orbital AI infrastructure by late 2027, though large-scale deployment depends on Starship’s launch capabilities and lower costs.

Skeptics point to major hurdles. An April assessment by the Government Accountability Office found that computing hardware does not cool down easily in space, that radiation can damage equipment, and that many more satellites would raise the risk of collisions. Launch, replacement and maintenance costs, and reliable data transfer between orbit and Earth, remain open problems.

In plain words: Musk says SpaceX hopes to start launching its space computers in the summer or fall of 2027, but he called that a guess, and he expects large numbers only in 2028 and 2029.

JBizNews Desk | Washington

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

New Tesla cars outside the company factory

On:

Archive photo: Steve Jurvetson / CC BY 2.0

SpaceX and Tesla have lost more than $420 billion in combined market value since Tuesday’s midterm elections, according to a report. A second report put the loss at about $456 billion between Tuesday’s close and Thursday’s close. Democrats won control of both the House and the Senate in the elections.

On Thursday, Tesla shares fell 10.77% to $343.69 and SpaceX shares fell 9.05% to $123.62, in a broader technology stock sell-off, according to the second report.

Why Washington matters to the stocks

Democratic control of Congress opens the prospect of more scrutiny of Elon Musk, who was among the largest donors to Republican candidates in the midterms. Most of the candidates he backed lost.

Deutsche Bank analysts wrote Wednesday: “While policy is naturally hard to predict, we could envision SpaceX facing some incremental headwinds in this type of backdrop.”

The Texas factor

Both companies are headquartered in Texas, where Democrats also won. Democrat James Talarico, a senator-elect, defeated Republican Attorney General Ken Paxton, and Democrat Gina Hinojosa, the governor-elect, beat Republican Gov. Greg Abbott. Both Republicans were longtime Musk allies.

The Deutsche Bank analysts noted that Hinojosa made data centers a major campaign issue. She called for a moratorium on new approvals until lawmakers adopt stronger protections on water use, electricity costs, local control and incentives, and she criticized tax breaks for large facilities.

Tesla and SpaceX are expanding artificial intelligence infrastructure in Texas and plan to build Terafab, a large complex that would include a data center and a semiconductor factory in Grimes County. SpaceX also plans a “supercomputer” at its Starbase launch site and company town, according to company job listings. The analysts said SpaceX’s Texas computing expansion could face delays, and that tighter data center rules could require the company to spend more.

In plain words: SpaceX and Tesla stocks have dropped sharply since the elections, and analysts are pointing to possible new scrutiny and rules in Washington and Texas.

JBizNews Desk | Austin, Texas

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Intel headquarters building in Santa Clara, California

On:

Photo: Coolcaesar, Wikimedia Commons, CC BY-SA 4.0. Image resized.

Intel will pay about $1.9 billion for Classiq, plus up to $350 million more if the Israeli quantum software maker meets specified milestones. The terms are in Intel’s quarterly report filed with the U.S. Securities and Exchange Commission for the period ended Sept. 26.

The extra $350 million is not guaranteed. Intel has not said what the milestones are or how much of that payment it expects to make. The filing says the $1.9 billion is subject to customary purchase-price adjustments, and that Intel can choose to pay in cash or in Intel shares. It has not said which it will use.

Intel announced Oct. 26 that it had agreed to acquire Classiq, a Tel Aviv-based maker of quantum computing software, in a deal meant to give the chipmaker a complete quantum computing product from chips to programs. The companies said the transaction is subject to regulatory approvals and customary closing conditions and is expected to close in the first half of 2027. Until then the two will operate separately.

What Classiq does

Quantum computers use the physics of very small particles to solve certain problems that ordinary computers handle slowly, and they come in several competing hardware designs. Classiq’s software lets developers describe what they want a quantum program to do and then builds the detailed instructions for whichever machine runs it, without requiring deep knowledge of quantum circuits. Intel compared that role to the design software that transformed conventional chip making decades ago.

Intel has invested in quantum computing since 2015 and says it is the only company that can produce both qubit chips, the basic units of a quantum computer, and the complex cryogenic control electronics at scale using standard semiconductor manufacturing. The combination would let Intel offer what it calls a full-stack quantum platform and add quantum processors to a lineup that already includes central, graphics and specialized chips.

What Intel says happens next

“Joining forces with Classiq transforms Intel’s Quantum Computing Group into a full-stack quantum computing business,” Chief Executive Lip-Bu Tan said in the release. Intel said it plans to keep supporting Classiq’s existing integrations with other platforms and to keep the software independent of any particular hardware, so that it keeps working with other makers’ quantum machines. Classiq co-founder and CEO Nir Minerbi said the two companies can speed up adoption of the technology.

In plain words: Intel will pay about $1.9 billion for Classiq, and could pay up to $350 million more if Classiq hits goals that have not been made public.

JBizNews Desk | Santa Clara and Tel Aviv

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Archive: Joshua Brown/Pexels (Pexels License). Cropped and resized. General archive, not a photo of the launch.

Amazon shoppers invited to a new service can now text Alexa to compare products, build a cart and buy without opening the shopping app. The rollout began Friday with a small group of U.S. customers who get an invitation by text, and Amazon said access will expand over the coming weeks.

If you have not received an invitation, you do not have it yet. Amazon said customers can sign up on a landing page to hear when more invites go out.

How it works

Invited customers accept the terms, link a phone number and start texting. The contact shows up on the phone as “Alexa Shopping.” Amazon said customers can text things like a weekly grocery order, a birthday gift search, or a photo of an outfit to find similar items.

The service can also compare products, show up to a year of price history, send an alert when a price drops, and give order updates. A reorder request is bought with the customer’s approval, Amazon said.

One conversation across devices

Amazon said a request started by text can carry over to the Alexa shopping tool in the Amazon app, on the website and on Alexa devices, and the other way around. The company said Alexa for Shopping will remember what you were working on and what you prefer.

In plain words: this is a way to shop by sending a text message, and for now only people Amazon invites can use it.

JBizNews Desk | Seattle
© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Archive: panumas nikhomkhai/Pexels (Pexels License). Cropped and resized. General archive, not a photo of the data center discussed.

Communities that voted to restrict data centers can keep those limits, White House economic adviser Kevin Hassett said Friday, while arguing planned AI construction elsewhere in the country can continue.

Data centers are the large buildings full of computers that run AI. How fast they get built can affect AI companies, local electricity bills and jobs in the towns where they are planned.

What Hassett said

Hassett, the director of the National Economic Council, said: “We respect the localities that decide that’s what they want to do. But America is a big place, and right now, the data centers scheduled over the next couple of years aren’t affected by these ballot initiatives.”

He added that there is “plenty of room in the U.S. for data center growth to continue, as it was projected ahead of the election.” He also said local communities should decide whether to welcome development.

Related coverage

Also Friday, President Trump said he is “100% in favor of Super Intelligence”. Separately, OpenAI’s finance chief said the company will keep building despite the election backlash.

JBizNews Desk | Washington
© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Photo: A large data center campus in Council Bluffs, Iowa / Chad Davis, Wikimedia Commons / CC BY 2.0

Communities hosting AI facilities and utilities planning new power capacity remain in OpenAI’s expansion plans despite this week’s election backlash, its finance chief said Friday. The company said it will keep investing and raising capital, and promised to cover the electricity and infrastructure costs of future campuses it owns and operates.

What the CFO said

Chief Financial Officer Sarah Friar said in an interview Friday that the midterm results do not change the company’s building plans. “When I think about the midterms, I need to keep thinking about how do we keep investing to meet that customer demand,” she said at the Citi TMT Leadership Summit in Pebble Beach, California.

“If the midterm means more focus on data centers and communities, we’re going to keep doing the work to get into those communities and to build trust with them,” Friar said.

The commitments

On Thursday, OpenAI published a National Community Compact for future campuses it owns and operates. The company pledged to pay the full cost of the electricity it uses and the grid infrastructure needed to serve it, provide financial assurances so other customers are protected if a project is delayed, reduced or canceled, publish site-level water and energy use, and give residents early input on project plans. It also said it will not use non-disclosure agreements with public officials, with narrow exceptions for security information and trade secrets.

Alongside the commitments, the company published a policy blueprint asking federal, state and local lawmakers to write similar rules into law for the whole industry. The blueprint is a proposal, not enacted law. In Georgia, OpenAI said its agreement with Georgia Power already lets the utility cut up to 1,000 megawatts of power to its planned Effingham County campus when demand peaks.

The announcement builds on pledges OpenAI made starting in January for its Stargate project and later in Georgia and Ohio.

Fundraising round and growth

In the interview Friday, Friar rejected reports that OpenAI’s current fundraising round is being done at a lower valuation or for a smaller amount than planned. The round is ongoing, she said.

She said OpenAI’s business grew 70% quarter on quarter in the third quarter, and that its enterprise business doubled over the same period. Those are the company’s own figures. “It’s saying people are actually using AI and getting value from it,” she said.

JBizNews Desk | San Francisco

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

The stock rout of the past 24 hours may have wiped billions from the net worths of the world’s wealthiest people. Still, the hiccup will do little to dissuade economists from believing that artificial intelligence will be a major driver of wealth gains in the coming years.

And while proposals for a billionaire’s tax have proved divisive in states like California, Microsoft co-founder Bill Gates suggests a similar policy is likely the best way forward in the age of AI.

Already, the fortunes of individuals who have founded AI or AI-adjacent companies have boomed, and the wealth effects for families invested in those companies promise to be significant.

But with massive wealth gains comes the question of inequality, and particularly in the case of AI, whether it might further exacerbate the current K-shaped economy. The K-shaped economy refers to the divergence between higher and lower earners and their economic prospects.

Philanthropic initiatives from the wealthy cannot be relied upon to address this gap, Gates said in a recent interview with The Financial Times. Rather, policymakers must move to tax wealth progressively, and then individuals can choose to donate accordingly from their earnings.

“The primary responsibility of dealing with inequality will always rest with governments,” Gates said. “We can’t expect philanthropy, which is voluntary in nature, to be the safety net.”

Gates, the founder of the Gates Foundation, decided to give “virtually all my wealth”—about $100 billion—to the organization early last year. The only catch was that it all had to be spent within the next 20 years.

Yet even these sums won’t go far enough to offset the inequality AI is likely to cause. According to Fed data, the top 1% of the wealthiest people in the U.S. own more than 50% of the nation’s stock market, meaning the gains experienced this year—and in the future—disproportionately flow to them.

“So, in a world of AI, there will have to be changes in tax systems, including the ability to tax people with huge wealth, the ability to tax billionaires,” added Gates.

“But whatever’s left over after those new taxes are put in place, I do think we’re seeing a very enlightened group that in both areas, here in the United States and for global good, they’re at a very young age already asking the questions about, ‘OK, what’s the most impact I can have?’ and ‘How do we be complementary to whatever government aid there is?’”

Gates’s other suggestion

Gates’s suggestion to tax the super-rich is—unsurprisingly—not universally popular among his tech-founder peers. Google co-founder Sergey Brin and former CEO Eric Schmidt have both flocked to tax-friendly states amid California’s proposed billionaire tax. Other entrepreneurs, like Nvidia founder Jensen Huang, say they understand why they may need to be taxed more.

But where Huang and Gates disagree is on the latter’s suggestion that robots should be taxed. In an essay shared in August, the entrepreneur and philanthropist wrote: “Right now, if you’re an employer and you hire someone, you pay payroll taxes on their earnings. But if you buy a robot, you can usually write it off right away as a business expense. The tax system nudges you toward replacing people with machines.”

As a result, Gates suggests taxing AI tokens and robots.

Huang disagrees. In an interview with Fox Business’s The Claman Countdown, the chipmaker boss said a few days later: “I love the heck out of Bill … but I don’t see what he sees. I see something very, very different. And so my remedies will be a little different.”

Huang added: “I’m in favor of taxes. And I think that … for anybody who is productive, it’s a great way for us to contribute back to society and the economy. But the fact of the matter is, there are probably lots of different ways to approach this.”

This story was originally featured on Fortune.com

This post was originally published here. 

jfrog-code-screen-generic

On:

Photo: Stanislav Kondratiev / Pexels

JFrog raised its full-year outlook Thursday after third-quarter revenue grew 31%, with cloud sales and spending by large enterprise customers driving growth at the software supply-chain company.

Revenue reached $179.2 million for the quarter ended September 30, while adjusted earnings were 32 cents a share, the company’s November 5 results announcement shows. GAAP earnings were 5 cents a share.

Shares rose about 35% to around $100.50 in regular trading Friday afternoon, after touching an intraday high of $101.90. The stock had jumped more than 20% in after-hours trading Thursday following the results. Friday’s figures are intraday, not closing prices.

Cloud business accounts for most revenue

Cloud revenue increased 60% from a year earlier to $101.5 million, representing 57% of total revenue compared with 46% a year earlier. The number of customers with annual recurring revenue above $1 million rose to 112 from 71, an increase of 58%.

Trailing four-quarter net dollar retention reached 124%, up from 118%. The measure reflects how spending by an existing group of customers changes over time, including expansions and reductions.

JFrog’s platform helps companies manage, secure and distribute software components and AI assets. For businesses buying those tools, the results show rising spending on the infrastructure used to control software delivery as developers and AI systems produce more code.

“AI agents are writing code faster than any developer can review it. Every output becomes an artifact that must be stored, secured, and governed,” CEO and co-founder Shlomi Ben Haim said in the company’s release.

Full-year outlook rises

JFrog now expects full-year revenue of $680 million to $682 million and adjusted earnings of $1.14 to $1.16 a share, JFrog’s earnings release shows. The previous forecasts were $648 million to $652 million in revenue and adjusted earnings of 96 cents to $1 a share.

For the fourth quarter, JFrog forecasts revenue of $183 million to $185 million and adjusted earnings of 28 cents to 30 cents a share.

The company generated $69.3 million in operating cash flow and $66.5 million in free cash flow during the third quarter. Cash, cash equivalents and investments totaled $880.9 million at September 30.

JBizNews Desk | Israel Business & Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Smartphone calculator and dollar bills on budget papers

On:

Archive: Tima Miroshnichenko/Pexels

Anthropic’s Claude chatbot can now look at your bank and card accounts and answer questions about your money in plain English. It is a test version, and it is rolling out this week to US customers on the paid Pro, Max and Ultra plans.

What it does

  • Shows your balances, your spending and your bills in one place.
  • Answers questions like “Where did my money go this month?” or “What’s left before payday?”
  • Points out charges that hit twice, surprise fees and bills that crept up over the year.
  • Explains money terms, such as what an APR is.

What it can’t do

  • It is read-only. It cannot move your money, pay a bill or invest for you.
  • It is not designed to give investment advice.
  • It can get things wrong. Anthropic says to ask which transactions are behind a number and to check important figures with your bank.

How your accounts get connected

Nothing connects unless you choose it. You sign in to your bank yourself through Plaid, a service that links apps to banks, and approve read-only access. Anthropic says Claude never asks for or sees your bank password.

For the accounts you pick, Claude sees the bank name, account name, type and last four digits, balances, credit limits and up to two years of transactions. Anthropic says chats in the Personal Finance section are not used to train its models. Chats you have elsewhere in Claude follow your regular privacy settings. You can unlink an account any time.

What it means for your wallet

If you already check several banking apps to see where your money goes, this puts it in one chat. The most useful part may be the catch-all question: “Does anything look off?” A double charge or a bill that quietly went up is easy to miss on your own.

Before you connect anything, think about what you are comfortable sharing, and keep your bank’s own app as the final word on your balance.

Anthropic says it is exploring ways for Claude to work alongside licensed professionals and regulated financial firms on bigger money decisions.

JBizNews Desk | San Francisco

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Aerial view of wastewater treatment tanks. Generic archive photo, not a New Jersey data center.

On:

Archive: Tom Fisk/Pexels

The New Jersey Assembly passed a measure on Oct. 29 directing the state to study how large-scale data centers use New Jersey’s water – and what that could mean for drinking water systems, wastewater infrastructure and the utility bills residents pay.

The bill, A3966, requires the Department of Environmental Protection to evaluate the short-term, long-term and anticipated future effects of water use by large-scale data centers on drinking water systems, wastewater systems and the environment. Under the measure, the DEP could request information from data center operators, public water systems and wastewater utilities, and the environmental protection commissioner could recommend new legislation on data centers’ water consumption.

Many data centers, beyond their heavy electricity use, rely on large amounts of water for cooling. The water study would run parallel to an ongoing Board of Public Utilities examination of data centers’ electricity use, giving the state a fuller picture of what the fast-growing facilities demand from local systems.

“In order to plan accordingly for future data center development, we need a clearer understanding of how these facilities use our water resources and what that means for our communities,” said Assemblyman Robert Karabinchak (D-Middlesex), a sponsor, in a Nov. 5 statement. “This study would give us the information needed to make informed decisions about how to protect our water supply and our residents.”

“As large-scale data centers require substantial amounts of water usage, New Jersey needs to understand the actual impact on our communities and water infrastructure,” said Assemblyman Ravi Bhalla (D-Hudson). “This study will help us better understand the resources these facilities require to function, so families don’t bear the cost of their consumption.”

“Water is an essential resource, and we need to understand how increased demand from data centers could affect our residents and communities,” said Assemblywoman Alixon Collazos-Gill (D-Essex, Passaic). “This legislation helps ensure water-use receives the same level of attention as electricity consumption.”

The measure now heads to the state Senate.

JBizNews Desk | Trenton

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

US Capitol and Reflecting Pool. Archive photo.

On:

Archive: Ramaz Bluashvili/Pexels

Democrats are preparing a wave of bills aimed at artificial intelligence and data-center development for when they take control of Congress in January, shifting from promoting AI growth to regulating it, Reuters reported.

The plans are proposals, not law. Among them: shielding consumers from higher electricity bills tied to power-hungry data centers, requiring independent testing of advanced AI models against safety standards, and clarifying that companies can be held liable when their AI systems cause harm.

Senator Chuck Schumer of New York, who is expected to become majority leader, said protecting consumers from energy costs linked to the data-center buildout would be a priority next year. Representative Lori Trahan of Massachusetts plans legislation requiring independent experts to evaluate leading AI models, her office said, along with a measure clarifying company responsibility when AI acts in harmful ways. Representative Zoe Lofgren of California said Democrats would pursue “responsible regulations for AI while also fostering innovation.”

The shift follows Tuesday’s midterm elections, in which voters backed candidates and measures skeptical of rapid data-center expansion and the strain it puts on local power grids. Technology shares sold off Thursday in part on concern that a new Congress will slow AI development.

Any bill would still need President Trump’s signature or a two-thirds majority in both chambers to override a veto. Analysts say Democrats could also attach AI measures to must-pass legislation. Most U.S. voters think the administration and Congress are not taking AI risks seriously enough, a Reuters poll found in October.

JBizNews Desk | Washington

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Stacks of US dollar bills. Generic archive photo.

On:

Archive: Adrian Vieriu/Pexels

America’s biggest companies have now borrowed more money through investment-grade bonds this year than ever before, and the giant technology firms building artificial intelligence are a big reason why.

U.S. investment-grade corporate bond sales reached a record $1.88 trillion through November 4, up 18.7% from a year earlier and topping the previous high of $1.86 trillion set in 2020, when companies stockpiled cash during the pandemic, LSEG data show.

Amazon and Alphabet led the borrowing. Together with Meta, Microsoft and Oracle, the big tech firms raised more than $370.7 billion, about a fifth of all such borrowing this year, largely to pay for AI data centers and computing power. Apple is the latest to join in, seeking to raise about $7.5 billion in a seven-part bond offering, according to a term sheet for the deal. That deal is not yet completed.

Notably, companies are selling fewer but much bigger bonds: the number of offerings fell about 4% to 1,084, so the average deal swelled to roughly $1.7 billion from about $1.4 billion a year ago.

Why it matters to your wallet

When companies borrow, they have to pay interest, and that cost ripples outward. Heavy corporate borrowing can push up the interest rates that shape mortgages, car loans and credit cards. It also affects retirement savings, since bond funds in 401(k) plans are buying much of this debt. Goldman Sachs forecasts that more than a third of big AI investment could be bond-financed by 2027, meaning even more borrowing is likely coming.

“The market has absorbed the AI issuance and it is evident from the oversubscribed nature of the deals,” said Victoria Fernandez, chief market strategist and fixed income portfolio manager at Crossmark Global Investments. “The AI companies do not seem to be slowing down in coming to both the USD and Euro bond markets.”

The open question is whether the enormous AI spending will generate profits fast enough to cover the mounting interest bills, especially with borrowing costs rising.

JBizNews Desk | New York

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Archive: Brett Sayles/Pexels

President Donald Trump on Friday threw his full weight behind the artificial intelligence buildout, declaring himself “100% in favor of Super Intelligence” and promising to work with states that want data centers and the riches he says they bring.

“As President, I am 100% in favor of Super Intelligence (SI), and will work with the many States in the U.S. that want to become RICH by having them,” Trump wrote in a 7:01 a.m. Truth Social post. “They are GOLD, and only foolish leadership will say ‘NO’ to the riches they bring!”

The statement lands days after voters in Tuesday’s midterms delivered the first broad electoral check on the AI industry’s expansion, backing candidates and measures in several states that promised tighter scrutiny of data center construction. It signals the White House intends to court willing states rather than retreat – while leaving the fights over where centers get built, and who pays for the power and water, at the state and local level.

What it means for your bills

The debate is ultimately about household costs. Communities near proposed sites have raised alarms that the massive facilities, which can use as much electricity as small cities, will push up power bills and strain local water supplies used for cooling. Supporters point to construction jobs, tax revenue and new investment.

Trump has tried to answer the cost concerns directly. In an October 21 post, he said large data centers should generate their own electricity, sell excess power to local utilities and deliver financial benefits to nearby residents. “MAKE DEALS, and get them built!!! GREAT FOR THE U.S.A.,” he wrote then.

Where the projects land now depends heavily on governors and statehouses, several of which changed hands Tuesday. Texas, the center of the construction boom, elected Democrat Gina Hinojosa governor on a platform that included pausing new data centers until lawmakers address electricity and water-supply concerns. Democrats also won governor’s races in Ohio, Georgia and Arizona, other states expected to attract major AI investment.

Read: Data center industry braces for state-level reckoning after midterm losses

JBizNews Desk | Washington

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Modern office towers in Tel Aviv, Israel. General city skyline, not a Microsoft campus.

On:

Photo: Jeffry Surianto / Pexels

Israel’s growing isolation could endanger its leadership in artificial intelligence and technology, Microsoft Israel R&D chief Michal Braverman-Blumenstyk warned in an interview with Ynet published Friday, saying the industry depends on global academic, customer and business links.

Braverman-Blumenstyk said Microsoft remains committed to Israel. The R&D center’s staff has roughly doubled to about 3,000 during her tenure, spread across some 30 product groups.

She steps down in December, a move announced last month. Ohad Yassin will replace her as head of the R&D center, which operates separately from Microsoft’s Israeli sales branch.

JBizNews Desk | Tel Aviv

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Photo: SpaceX / CC0

On:

Photo: SpaceX / CC0

Scottish Mortgage Investment Trust, one of Elon Musk’s longest-standing backers, said Friday that its SpaceX stake has fallen to 11.1% of total assets, down from 15% at the end of September.

The Baillie Gifford-managed fund trimmed the position after lockup restrictions from SpaceX’s blockbuster June IPO lifted, and said it keeps high conviction in the rocket and satellite company. The position added 8 percentage points to returns and helped drive a 31.5% net asset value gain in the six months through September 30, against a 16% return for the FTSE All-World Index, the trust’s half-year results show. Proceeds also helped the trust repay borrowings and reduce gearing.

In the same report, the trust revealed that it sold the last of its Tesla shares back in April – months before this week’s slide in the stock. The sale ended an investment dating to 2013 that generated about $6.2 billion in realized profit on roughly $470 million invested – a 13-fold return over 13 years, CNBC reported.

Scottish Mortgage first invested in SpaceX in December 2018, when the company had a reported valuation of about $31 billion. The trust invests globally in public and private high-growth companies with a long-term view.

JBizNews Desk | London

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Photo: Alexander Kubitza / Public domain

On:

Photo: Alexander Kubitza / Public domain

Axon is putting $1.2 billion into Skydio, the largest U.S. maker of police drones, in a deal that makes the Taser company the exclusive public-safety distributor of Skydio products in the United States and Canada.

The investment, in cash and stock, adds about 20% to Axon’s fully diluted ownership of Skydio at a $5.2 billion pre-money valuation, the company said. The deal is not done: closings are expected in the fourth quarter of 2026 and the first quarter of 2027.

Skydio serves more than 1,500 public-safety customers. Its drones are used by police departments for “drone as first responder” programs, where a drone launches ahead of officers to give them a live view of a scene. Axon says pairing its body cameras, Tasers and Axon Evidence software with Skydio’s drones will speed up emergency response times – a claim made by the company, not yet independently measured.

The investment lands as Axon trims its own outlook. The Scottsdale, Arizona company posted third-quarter revenue of $908.6 million, up 28% from a year earlier, but cut its 2026 growth outlook to about 30% – $3.61 billion to $3.63 billion in revenue – from 32% to 34%. Shares fell about 14% in extended trading Thursday after the report. The stock fell nearly 9% further in premarket trading Friday as investors weighed the weak fourth-quarter guidance and the Skydio investment, CNBC reported.

Skydio, based in San Mateo, California, has become the dominant American drone supplier to government agencies as U.S. restrictions on Chinese-made drones have pushed departments away from DJI, the Chinese market leader. The Axon deal deepens that position by putting Skydio’s products inside Axon’s vast public-safety sales channel.

JBizNews Desk | Scottsdale, Arizona

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Empty cinema with red seats and a blank screen.

On:

Photo: Tima Miroshnichenko/Pexels

The New York Times Company is in exclusive talks to acquire Letterboxd for roughly $330 million, The New York Times reported Thursday, citing two people familiar with the confidential discussions. No agreement has been reached, and the talks could still fall apart, the report said.

The proposed price would be more than 20 times Letterboxd’s projected 2026 earnings of about $15 million, according to the Times report. The Times Company declined to comment, and Letterboxd co-founder Matthew Buchanan did not respond to the newspaper’s request for comment.

A film community with 30 million users

Letterboxd has an audience of about 30 million movie fans, the Times reported. Founded in 2011, it gives people a place to record what they watch, review and rate films, follow friends and build watchlists. Its official description also lists paid features such as an ad-free experience, viewing statistics and streaming-service filters.

The Times reported that Letterboxd’s parent company has been working with investment bank LionTree on a sale. A24 and Sony were among the prospective buyers drawn to the service’s engaged audience, according to the report.

Founders could stay, with a standalone service

The discussions have included keeping Buchanan and fellow co-founder Karl von Randow in charge, reporting to a Times Company executive, the newspaper reported, citing one person familiar with the talks. Letterboxd would most likely remain a standalone site for the foreseeable future, another person told the Times.

The report said the companies had discussed connecting the Times Company’s advertising technology with Letterboxd’s systems. Both businesses draw revenue from advertising and subscriptions.

Expanding the bundle beyond news

A purchase would extend the Times Company’s effort to build products beyond its main news service. The Times report placed the potential acquisition alongside the company’s earlier purchases of The Athletic, Wirecutter and Wordle. It said the company had 13.8 million subscribers and was aiming for 15 million by the end of 2027.

JBizNews Desk | Media & Technology

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Photo: Firefly Aerospace

On:

Photo: Firefly Aerospace

AST SpaceMobile said Friday it will build its own reusable heavy-lift rocket, named Blacklift, designed to carry up to five of its BlueBird satellites to low Earth orbit per launch starting in late 2028, using engines supplied and licensed by Firefly Aerospace under agreements valued at more than $1.3 billion.

The rocket completes AST SpaceMobile’s long-held strategy of end-to-end vertical integration, pairing the launch vehicle with the company’s in-house production of the largest phased-array satellites operating in low Earth orbit. BlueBird satellites are now in production through BlueBird 51 at the company’s facilities in Midland, Texas. AST SpaceMobile said it will keep using outside launch providers as well, with agreements now covering more than 100 BlueBird satellites to be deployed cumulatively between 2027 and 2028.

Under the license agreement, AST SpaceMobile will pay approximately $800 million in upfront consideration: $400 million in cash plus 10 million to 14 million shares of Class A common stock, with the share count set by a forward VWAP calculation. A separate supply and services agreement covers $250 million for the delivery of engines for the first three rockets and $250 million over three years for engineering support. The $1.3 billion-plus total combines cash, stock and services; it is not an all-cash payment.

Firefly will deliver 27 of its Miranda engines and nine of its vacuum-variant Vira engines for the first three Blacklift rockets, and license the engine intellectual property so AST SpaceMobile can manufacture the engines at a new U.S. facility. Firefly keeps full ownership of the IP and will continue serving the broader government and commercial launch market. The Miranda engine, a 230,000 lbf-thrust engine scaled from Firefly’s flight-proven Reaver and Lightning engines, has completed more than 175 hot-fire tests totaling over 4,200 seconds.

“Firefly is one of a few companies in the United States with a proven liquid engine of this scale, making it a scarcity in the industry,” said Abel Avellan, founder, chairman and CEO of AST SpaceMobile. Firefly CEO Jason Kim called it “a bold transformational deal” for both companies, adding that AST SpaceMobile gets control of its own launch schedule while Firefly gains funds to expand rocket production, including new test bays at its Rocket Ranch in Briggs, Texas.

Scott Wisniewski, AST SpaceMobile’s president, said the Miranda engine lets the company “leapfrog the traditional five-to-ten-year development cycle” and expand its addressable market beyond direct-to-device service into U.S. government radar, secure communications, IoT, federal emergency and AI edge-compute work. AST SpaceMobile will host a Blacklift unveiling event in New York City on Thursday, November 12, webcast on its investor relations site.

JBizNews Desk | Midland, Texas

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Apple Store glass facade with a white Apple logo at night.

On:

Apple has an undisclosed multilingual gray list of tens of thousands of words and phrases that its iOS keyboard will not autocomplete or swipe-type, a researcher said on Friday.

The list is not a ban. iPhone users can still type what they like when drafting emails and texts. But the words on the list will not be suggested to users via autocomplete or swipe-typing in most cases, opening a window into how Apple manages the tradeoff between preventing its devices from offering offensive or politically charged suggestions to users and anticipating what they are trying to say.

Hundreds of millions of people worldwide use iPhones, so Apple’s choices about which words its keyboard suggests could quietly shape how they communicate. Apple has not publicly disclosed the list, the researcher said.

The flagged words and phrases are overwhelmingly vulgar or offensive or related to drugs, sex or violence, according to independent researcher Derik Kauffman, who discovered the list recently. But other phrases are just politically salient, like “dictator Trump” and “sleepy Joe.”

In an email, Apple did not comment directly on Kauffman’s findings, including the size or composition of its list, or its reasons for flagging certain political terms. It said that iPhone users always had the option to type specific words or phrases and could always override any predictive text suggestions. Apple said its keyboard was set up to avoid offering suggestions that might be inappropriate or offensive.

Apple (credit: SHUTTERSTOCK)

Other phrases excluded from iOS’s predictive text features include “right wing conspiracy” as well as “Tibetan independence” and “United Ireland,” according to Kauffman. Non-English languages have different flagged words or phrases. In Hindi, for example, the keyboard does not allow users to swipe-type “Modi braggart” or “Rahul is Pappu,” digs at India’s Prime Minister Narendra Modi and his leading opponent, Rahul Gandhi, respectively, according to Kauffman.

He said French speakers could not swipe-type the French phrase “immigration too high” and that Hebrew speakers could not swipe-type “occupied territories.”

“Every major language has some version of this,” Kauffman said.

Kauffman said it was understandable that Apple wanted to keep users from accidentally sending obscene messages or inflammatory slogans. But he said there were inconsistencies in how Apple applied the feature. For example, he said “read smut” could be swipe-typed while “read gay smut” could not.

“It’s a very light-touch form of censorship,” Kauffman said. “But at the scale that Apple is operating, it only takes a light touch to have a big impact.”

How the gray list was discovered 

Kauffman said he found the list in September after noticing that his swipe-typing – the practice of gliding a finger back and forth over a keyboard rather than pecking out each word individually – would stutter over certain phrases, like “bad boy” or “white man,” but not “bad girl” or “black man.”

His curiosity piqued, Kauffman said he used a combination of OpenAI’s Codex and Anthropic’s Claude Code to hunt for a reason why, eventually stumbling on a compendium stored “deep in the bowels” of Apple’s operating system. He said there were different ways to count the number of gray-listed words and phrases and their various permutations, but a conservative estimate would put them at about 100,000, not including Chinese characters.

Reuters tested a small subset of the phrases identified by Kauffman on an iPhone and found they were all difficult to enter through swipe-typing or were not suggested via autocomplete. Attempts to swipe-type “United Ireland” or “Tibetan independence” often yielded “United Iceland” or “Tibetan independences,” for example. Trying to swipe-type “white man” would yield “white mean,” while “red man” or “blue man” could be swipe-typed without difficulty. “Sleepy Joe” would yield “Sleepy hope.” “Dictator Trump” could be swipe-typed, but “trump” was lowercase and the phrase was not suggested in autocomplete – the keyboard instead suggested, “dictator trumps.” Swipe-typing “read gay smut” would yield “Read gay Smit.” “Read smut” could be swipe-typed with no issue.

Matt Suiche, an engineer at agentic AI cybersecurity company Tolmo, said he was able to independently recover the same gray list and cross-reference Kauffman’s findings against Apple’s internal data. He said he got a near-perfect match.

Suiche said the Apple list had been hidden in plain sight for years, providing an interesting look at how “human communication has been potentially influenced and filtered for years inside messaging apps.”

A call for transparency 

Kauffman is not alone in having picked up on the nudges from iOS’s predictive text feature. Over the years, social media users have wondered why they could not swipe-type certain phrases. Earlier this week, a Norwegian systems administrator independently discovered a list of words on Apple’s gray list, including “al-Qaeda,” “Auschwitz” and “Hooters,” but the existence of a wider list of sensitive phrases is reported here for the first time.

Apple, like other tech companies, has previously run into trouble when predictive technology suggests embarrassing or controversial texts to users. Alphabet-owned Google’s autocomplete feature has regularly attracted criticism – and sometimes even litigation – over its suggested search terms. Last year, Apple drew media attention when users reported that dictating the word “racist” to their iPhones would briefly cause the word “Trump” to flash across their screens. Apple said at the time that the problem was an error and had been fixed.

David Greene, the senior counsel at the Electronic Frontier Foundation, a digital civil liberties group, said that while a part of him found the word list “creepy” and he wondered about Apple’s judgment calls, he could see an argument for it being a useful way to curb unwanted or inflammatory suggestions.

“But if they are going to do it, this is something users should be really well aware of and have some ability to customize,” Greene said. “This is something that should not be secret.”

This post was originally published on here. 

Generic cold-storage racks

On:

Photo: Anna Shvets / Pexels

WASHINGTON, Nov. 5, 2026 – U.S. agencies warned Thursday that malicious cyber activity is targeting industrial refrigeration and cold-chain monitoring systems globally, putting temperature-sensitive operations at grocers, healthcare facilities and transport businesses at potential risk.

The joint advisory from CISA, the FBI and the departments of Health and Human Services, Agriculture and Transportation urges owners to remove unnecessary internet access, secure remote connections and update the systems that control and monitor refrigeration.

The agencies described possible consequences of successful attacks, including temperature changes, inventory loss, equipment damage and fire hazards.

Why refrigeration controls matter

The systems are used for automated temperature monitoring, continuous temperature records, defrost management and energy use. They are common in healthcare and public health, chemicals, food and agriculture, and transportation, including refrigerated transport services, the advisory said.

That makes the issue broader than a retailer’s computer system. A disruption to equipment controls or monitoring can affect the physical conditions in which temperature-sensitive inventory is held and moved.

The agencies warned that attackers could alter temperature settings or prevent operators from seeing current temperatures. The latter can make it harder for staff to identify a problem even when they rely on automated monitoring.

Possible physical effects, not confirmed loss totals

Unsafe temperature changes could reduce equipment life, create a fire hazard or cause inventory losses, the advisory said. It also warned of possible indirect harm to consumers if products were consumed after unsafe storage.

CISA official Chris Butera said every organization dependent on refrigeration and monitoring should assume it could be a target. FBI official Brett Leatherman said systems were being targeted across critical-infrastructure sectors globally.

What owners and operators can do

The agencies’ first recommendation is to eliminate public internet exposure for control equipment. Where remote access is necessary, they advise protecting it through appropriately secured and updated network equipment rather than leaving controller panels openly reachable.

They also recommend separating operational equipment from business networks, changing default passwords, using strong passwords and deploying phishing-resistant multifactor authentication wherever feasible. Access to make changes should be limited to authorized devices and users.

Keeping firmware current is another priority. Owners should identify equipment that is no longer supported by its manufacturer and plan to retire or replace it, rather than assuming an old controller can be made safe through a password change alone.

Reliable backups should be kept offline and offsite and tested, the advisory said. Operators should also enable and review logs for unexpected activity, define safe temperature-setting limits and establish alerts when equipment moves outside expected ranges.

A business-continuity issue across the cold chain

For grocery operators, cold-storage companies, healthcare businesses and refrigerated transport providers, the immediate task is to know which controls are exposed, who can change them and how staff will detect an unsafe change.

The advice is aimed at system owners and operators.

The agencies urge organizations to report suspicious or criminal activity to CISA, the FBI’s Internet Crime Complaint Center or a local FBI field office. Healthcare operators facing possible medical-supply-chain disruption can also seek assistance from HHS.

JBizNews Desk | Washington

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

AI21 Labs logo

On:

Photo: AI21 Labs / Wikimedia Commons

AI21 Labs has shifted its focus to training and optimizing AI agents for enterprise customers after cutting about 110 of its roughly 180 employees in May, founders Ori Goshen and Yoav Shoham said in an interview with Globes published Friday. About 70 employees remain.

The founders denied reports that the company raised $300 million shortly before the layoffs, and denied that it had sought a sale to Nvidia, which endorsed the company last month.

Goshen and Shoham declined to comment on reports of acquisition talks with Nebius, confirming only a commercial partnership with the company.

Co-founder Amnon Shashua left AI21 in June, a month after the layoffs.

The founders said the company’s approach lowers the cost of running AI agents and improves their performance, claims that are the company’s own.

JBizNews Desk | Tel Aviv

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

On:

Photo: BalticServers / Wikimedia Commons

Data centers emerged from the midterm elections facing a tougher regulatory map, as Democrats who campaigned on clamping down on the energy-hungry facilities swept governors’ races and statehouses across the country.

Data center regulation was not on most ballots and was not a top issue for most voters. But hardline stances against the server warehouses were woven into many winning campaigns, often blended with affordability messaging over electricity prices. Campaigns spent more than $380 million on television ads mentioning data centers, with Democrats outspending Republicans more than three-to-one, according to AdImpact data.

Wisconsin Gov.-elect David Crowley, who hardened his position under campaign pressure, told the AP on Thursday that he is discussing proposals to require developers to disclose more about their water and energy consumption, to ensure a community supports a project before it can go forward, and to insulate ratepayers and taxpayers from projects’ energy costs. These remain proposals under discussion; no moratorium or new rules have been enacted.

Data Center Watch, a project of AI security consultancy 10a Labs, said the vast majority of winning candidates for Congress and governorships it tracked had expressed support for stronger regulation or even a moratorium.

Morgan Stanley told clients the decisive Democratic results carry “greater expectations for data center restrictions” that could “split the country into areas that are open to new data centers and those without.” The bank called Texas Gov.-elect Gina Hinojosa’s victory the “most important election result” for data center development and said Ohio Gov.-elect Amy Acton’s win “creates significant risk of data center project delays.” Hinojosa has said she would block a permit for any data center that was not a good deal for Texans or where residents had no say.

The industry is signaling concern. Data Center Coalition official Dan Diorio said developers expect newly elected leaders to try to deliver on moratorium pledges, warning that host communities could miss out on tax revenue and payments from tech companies, and that data centers are driving a broader economic expansion that encourages investment in domestic manufacturing.

OpenAI on Thursday published a framework of commitments to the communities where it builds data centers, asking policymakers to write “clear rules for how the industry builds” and saying lawmakers should “advance responsible investment, protect residents, and ensure that the benefits of AI infrastructure are broadly shared.”

The stakes extend to markets and growth. The tech-heavy Nasdaq sank 11% below its all-time high Thursday amid worries that data center construction may slow as borrowing gets costlier, and AI data center construction has been driving much of U.S. economic growth. Texas is poised to overtake northern Virginia as the largest global data center market by 2030, according to commercial real estate giant JLL.

Voters are broadly skeptical: about three-quarters of U.S. voters oppose the construction of a new data center in their community, according to the AP Voter Poll. Still, only about a quarter said AI data centers were crucial to their vote for Congress, well below the shares citing government corruption or the cost of living.

JBizNews Desk | Harrisburg

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Zcash logo on a black background

On:

Photo: Zcash / Wikimedia Commons

Zcash activated its NU7 network upgrade early Friday, cutting the target time between blocks from 75 seconds to 25 and retiring the privacy coin’s original Sprout pool, where about 12,224 ZEC (about $15 million) can no longer be spent.

The upgrade took effect at block 3,543,405 at 02:34 UTC Friday (9:34 p.m. ET Thursday). In its first hours the chain was still speeding up: 225 blocks had been mined by 04:59 UTC, an average of about 39 seconds apart, with mining difficulty still adjusting toward the new target.

Faster private payments

Zcash lets people send money while hiding the sender, recipient and amount. At the new target, a merchant could see a payment’s first confirmation in about 25 seconds on average, and an exchange requiring three confirmations could cut its expected wait from nearly four minutes to about 75 seconds.

NU7 also caps how much shielded activity fits in each block, reducing the worst-case data wallets must download each day from 271 megabytes to 169, while estimated capacity for simple private payments could more than double to over six per second.

Sprout funds locked

Under the ZIP 2003 specification, NU7 rejects version 4 transactions, the only format that can spend funds from Sprout, Zcash’s original 2016 shielded pool. The remaining coins cannot be transferred under the new rules but have not been burned.

The specification leaves open the possibility that a future network upgrade could restore access, while warning that the ability to spend Sprout funds “may never be re-enabled.” Co-founder Sean Bowe has outlined a possible recovery mechanism, though none had been approved as of Friday.

Fees set aside for future miners

Miners now receive 40% of transaction fees. The remaining 60% is removed from circulation and recorded in a reserve that starts feeding future block rewards around February 2031 under a Network Sustainability Mechanism, which preserves Zcash’s 21 million-coin supply cap and halving schedule.

JBizNews Desk | New York

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Equinix data center campus, file photo

On:

Photo: Equinix, Inc. / Wikimedia Commons

Equinix says its smaller data centers could be less exposed to potential U.S. construction restrictions as regulators scrutinize the rapid buildout of artificial-intelligence infrastructure, its Asia-Pacific president says.

The company’s facilities typically range from 30 to 60 megawatts, below the roughly 100-megawatt threshold that U.S. proposals generally target, Cyrus Adaggra, Equinix’s Asia-Pacific president, said. The assessment is the company’s own; the proposals and their thresholds are not settled law.

“We will invest in markets where we see customer demand,” Adaggra said, adding that the company’s footprint typically falls under the thresholds regulators have discussed.

Capex rises on customer demand

Equinix has increased capital spending from about $2 billion a few years ago to $7 billion, driven by customer demand, Adaggra said. Annualized gross bookings exceeded $500 million for the first time in the third quarter, up 35% from a year earlier. The company serves about 60% of Fortune 500 companies, he added.

Potential restrictions on U.S. data center development could raise costs across the AI infrastructure supply chain, and slower U.S. construction could weigh on demand for semiconductors and other equipment, including from Asian suppliers, according to Russell Investments.

JBizNews Desk | Ashburn

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.

Elon Musk, file photo

On:

Photo: Debbie Rowe / The Royal Society

Elon Musk’s paper wealth fell about $113 billion over two days as Tesla and SpaceX shares slid following Democrats’ midterm-election sweep, the Bloomberg Billionaires Index shows.

Musk lost about $33.5 billion on Wednesday and another $79.7 billion in Thursday’s selloff, as of the Nov. 5 market close. Tesla shares fell 3.3% and SpaceX shares fell 3.8% on Wednesday even as the Dow Jones Industrial Average posted its best day after a midterm election in 44 years. On Thursday the Dow gave back those gains, falling 1,833 points, about 3.5%, in a broader technology-stock selloff that took the S&P 500 down about 4.4% and the Nasdaq about 5.5%.

About $456 billion in company value erased

Combined, the two companies’ market value fell from about $3.5 trillion to about $3.04 trillion over the two days, a loss of roughly $456 billion, the companies’ closing values show. Tesla ended the two days at about $1.36 trillion in market value, about 86% of its level before the slide, and SpaceX at about $1.68 trillion, about 87%.

Investors are weighing the prospect of closer congressional scrutiny of Musk’s businesses after Democrats’ election gains. Democrats have signaled plans to investigate Musk’s role in the Department of Government Efficiency and potential conflicts of interest, as SpaceX holds billions of dollars in federal contracts. A spokesperson for Rep. Robert Garcia, the incoming House Oversight Committee chair, said last month that Musk’s DOGE role “is something we plan to investigate.”

“The nail that sticks out gets hammered first, and Musk has certainly stuck out,” Keith Snyder, senior equity analyst at CFRA Research, said. Snyder called the selloff a “knee-jerk reaction” to the election, noting that President Donald Trump retains veto power over major AI restrictions and remains friendly toward Musk.

Musk spent an estimated $427.5 million to $483 million backing Republicans in the midterms, the spending estimates show.

JBizNews Desk | New York

© JBizNews.com. All rights reserved. Unauthorized reproduction or redistribution is strictly prohibited.