Anthropic faces a proposed class-action suit accusing the company of overselling how much access customers get to its premium Claude plans — a dispute that could complicate the AI startup’s path to a public offering and sharpen interest in decentralized AI alternatives.
What the lawsuit says
- Filed Monday in the U.S. District Court for the Northern District of California, the complaint was brought by Washington, D.C. resident Karl Kahn, who is seeking class-action status for customers who paid for Anthropic’s higher-tier Claude subscriptions since April 2024.
- The dispute centers on Anthropic’s “Max 5x” and “Max 20x” plans (priced at roughly $100 and $200 per month). The company advertised these tiers as providing five- and twenty-times the usage of its standard Pro plan.
- Kahn’s suit alleges actual usage caps fall far short of those multipliers, are unpredictable, and force subscribers to curtail work, ration usage, or buy extra access. He says a single five-hour development session consumed about 15% of his weekly allowance on the Max 20x plan.
- The complaint cites July 2025 emails Anthropic allegedly sent to customers that outlined weekly allowances across models and tiers, arguing those disclosures show a gap between marketing claims and real-world access.
- The filing asks the court to find Anthropic’s marketing misleading or fraudulent and seeks relief on behalf of affected subscribers.
Why this matters now
- The lawsuit arrives while Anthropic is drawing investor attention and preparing for a possible IPO, increasing the stakes of any public legal or regulatory trouble.
- Legal scrutiny of AI firms is not unique to Anthropic: OpenAI recently faced a multistate probe tied to consumer harm claims around ChatGPT, which surfaced as the company explored its own IPO filing.
Context on access controls and the crypto angle
- This case intersects with broader concerns about centralized control over advanced AI. Days earlier Anthropic suspended access to its Fable 5 and Mythos 5 models after complying with a U.S. government directive tied to export controls; the company said the order required restricting access for foreign nationals, including employees abroad.
- Critics and crypto-aligned observers have used such incidents to argue for decentralized approaches to AI infrastructure. CoinFund founder Jake Brukhman noted that large models concentrate control in a few companies and pointed to decentralized training projects — including Gensyn, Prime Intellect, Bagel, Pluralis, Nous Research, Macrocosmos AI, and Covenant — as efforts to distribute compute and reduce single points of control. He suggested decentralized training could become a viable alternative, though technical hurdles remain.
Bottom line
The new lawsuit adds consumer-facing legal risk to Anthropic’s plate just as it navigates export-control fallout and IPO-related pressure. For the crypto and decentralized-AI communities, the case underscores two trends: users and regulators are scrutinizing how AI companies monetize and limit access, and decentralized projects may gain attention as responses to perceived centralization of compute and model access.
Read more AI-generated news on: undefined/news
