Amazon has quietly changed the legal rules governing millions of U.S. customers, bringing back mandatory arbitration and barring most consumers from joining class-action lawsuits against the company.
The new terms took effect immediately for customers who continue using Amazon’s services. Instead of taking most disputes to court, customers will generally be required to pursue claims individually through binding arbitration. Small-claims court remains available for eligible disputes.
The change matters because class actions allow large numbers of customers with similar complaints to combine their claims into one case. Without that option, a consumer alleging a relatively small financial loss may have to decide whether pursuing an individual claim is worth the time and effort.
Amazon says arbitration provides a faster and less expensive way to resolve disputes.
But the company has seen firsthand how expensive arbitration can become when customers organize at scale.
Amazon previously abandoned mandatory arbitration in 2021 after roughly 75,000 individual arbitration claims were filed over allegations involving Alexa recordings. Because companies can be responsible for substantial filing and administrative fees in arbitration, the wave of cases created a costly problem for Amazon.
The new rules appear designed to address that vulnerability as well.
Amazon now defines 25 or more similar claims filed within a six-month period as “mass arbitration.” Those cases can be processed in batches rather than all moving forward simultaneously.
That gives Amazon greater control over one of the strategies plaintiffs’ lawyers have increasingly used against companies with arbitration clauses: filing thousands of individual cases at once.
The implications extend beyond Amazon.
Many consumer businesses have spent years adding arbitration clauses and class-action waivers to contracts covering everything from credit cards and cellphone plans to ride-sharing apps and subscription services.
Amazon’s reversal could encourage other large companies to reconsider whether arbitration provides stronger protection from large consumer lawsuits.
For customers, however, the practical change is straightforward.
A dispute involving a damaged purchase, subscription, privacy allegation or another Amazon service may now be significantly harder to turn into a large collective lawsuit.
Customers can still bring legitimate claims.
They will simply be far more likely to have to do it one person at a time.
JBizNews Desk | Seattle
© JBizNews.com All Rights Reserved. Reproduction or distribution without written permission is prohibited.
Jane Street Absorbs $15 Billion AI-Related Hit — and Is Still Having a Record Year
Jane Street, one of the most powerful trading firms on Wall Street, suffered an extraordinary $15 billion hit in July after an AI-stock selloff battered positions connected to one of the market’s most aggressive artificial-intelligence investment funds.
Yet the loss reveals something equally remarkable: Jane Street has still generated more than $40 billion in trading revenue this year, already surpassing the $39.6 billion it produced during all of 2025.
The July setback was tied partly to Jane Street’s investment in Situational Awareness, an AI-focused hedge fund run by former OpenAI researcher Leopold Aschenbrenner.
The fund had grown rapidly as AI-related stocks surged during the first half of the year. But when semiconductor, memory and other AI-linked shares suddenly reversed in July, leveraged positions came under severe pressure.
Situational Awareness ultimately unloaded much of its stock portfolio in a distressed sale to Citadel after losses triggered margin calls.
Jane Street was caught in that reversal both through its investment in the fund and through other technology positions of its own.
Several major memory and semiconductor stocks fell roughly 50% during the July rout, according to a Jane Street communication to employees.
The result was Jane Street’s first negative month of trading revenue since 2016.
For perspective, a $15 billion loss would be catastrophic for almost any investment firm in the world.
For Jane Street, it interrupted an otherwise extraordinary year.
The privately held trading company has approximately 3,500 employees and operates across more than 200 trading venues worldwide, buying and selling stocks, bonds, ETFs, options, currencies and commodities.
Its scale allows the firm to hold enormous positions while providing liquidity to global markets.
That model can be extraordinarily profitable when markets move as expected.
July demonstrated what happens when they do not.
Jane Street said it has since reduced risk in some strategies and closed significant portions of positions associated with the losses.
The episode also offers investors a rare glimpse into how concentrated the AI trade has become.
Artificial intelligence is no longer simply a collection of popular technology stocks held by retail investors. Hedge funds, proprietary trading firms, banks and institutional investors have committed enormous amounts of capital to many of the same semiconductor, data-center, cloud-computing and memory companies.
That concentration can amplify gains when AI stocks rise.
It can also accelerate losses when investors attempt to exit similar positions simultaneously.
The most unusual part of Jane Street’s July loss may therefore be what happened afterward.
Despite absorbing approximately $15 billion in a single month, the firm remains on pace for what could still be the most profitable year in its history.
That says as much about the extraordinary amount of money being made around today’s markets as the loss itself.
But July delivered a warning that applies far beyond Jane Street:
A trade can become enormously profitable without becoming less dangerous.
And when billions of dollars are crowded into the same AI bets, a relatively short market reversal can produce losses measured not in millions — but in tens of billions.
JBizNews Desk | New York
© JBizNews.com All Rights Reserved. Reproduction or distribution without written permission is prohibited.
