Anthropic hit with proposed class-action over premium Claude usage limits
Anthropic is facing a proposed class-action lawsuit alleging that customers who paid up to $200 a month for its premium Claude plans have received far less usable capacity than the company’s marketing suggested.
What the suit says
- Plaintiff Karl Kahn, a Washington, D.C. resident, filed the complaint Monday in U.S. District Court for the Northern District of California seeking class status for customers who subscribed to Anthropic’s higher-tier Claude plans since April 2024.
- The target: Anthropic’s Max 5x and Max 20x subscriptions — $100 and $200 per month tiers that the company markets as providing fivefold and 20-fold usage over its standard Pro plan.
- The filing alleges actual usage caps and throttles are far lower than those advertised and difficult for subscribers to predict before hitting limits. Kahn says he upgraded to Max 20x for coding and development work, but a single five-hour session used roughly 15% of his weekly allowance, forcing users to pause work, ration usage, or buy extra access.
- The complaint cites July 2025 emails Anthropic allegedly sent to subscribers that outlined expected weekly usage allowances across models and tiers, arguing those disclosures show a meaningful gap between marketing and real access.
- The suit asks the court to rule Anthropic’s marketing was misleading or fraudulent and seeks relief for impacted subscribers.
Why it matters now
The case lands as Anthropic draws heightened investor attention ahead of a much-anticipated public offering, and comes amid other legal and regulatory headwinds for AI companies. OpenAI recently faced a multistate probe into alleged consumer harms linked to ChatGPT — a spotlight that intensified as OpenAI was reported to have confidentially filed IPO paperwork.
This legal pressure follows a separate operational controversy for Anthropic earlier this month: the company suspended access to its Fable 5 and Mythos 5 models after complying with a U.S. government export-control directive. Anthropic said the order required restrictions on foreign nationals — including employees located both inside and outside the U.S. — prompting the company to disable those models while keeping other Claude variants available.
Bigger picture: centralization vs. decentralized AI
The dispute also feeds broader debates about centralized control of advanced models. CoinFund founder Jake Brukhman recently argued that gated access to large-scale models highlights centralization risks and helps explain growing interest in decentralized AI training and compute networks. He pointed to projects such as Gensyn, Prime Intellect, Bagel, Pluralis, Nous Research, Macrocosmos AI, and Covenant as teams working on distributed training systems that, in time, could rival centralized providers — though he noted significant technical hurdles remain.
Bottom line
The lawsuit places Anthropic’s subscription business under fresh scrutiny alongside regulatory questions and model-access controversies. As Anthropic courts public markets and fights multiple fronts of criticism, how courts and regulators respond could shape both consumer expectations for paid AI tiers and the broader competition between centralized and decentralized AI infrastructures.
Read more AI-generated news on: undefined/news
