Anthropic Sued Over Claude's 'Misleading' $100-a-Month Subscription Limits

Key points

• A class-action lawsuit filed against Anthropic alleges its Claude Max subscription plan, priced at $100 or $200 per month, misleads customers about how much AI usage they actually receive.
• The lawsuit was brought by Monica Vaca and Kati Daffan, two attorneys who together spent 38 years at the US Federal Trade Commission, the government body that enforces consumer protection laws.
• Anthropic's Max plan advertises "5x" or "20x" the usage of its $20/month Pro plan, but the lawsuit claims the real limits are far smaller once hidden session rules apply.
• Anthropic introduced the disputed weekly usage limits in August 2025, several months after the Max plan launched in April 2025.
• Anthropic has not publicly responded to the expanded complaint.

Paying $100 a month for software you rely on to keep your job is serious money. Some Claude subscribers say they handed over that cash, sometimes $200 a month, expecting dramatically more room to work, then hit a wall almost immediately. Now they have lawyers.

What is the lawsuit actually claiming?

The lawsuit, first reported by The Verge AI, says Anthropic advertised its Max plan with big, simple numbers that don't tell the full story. The plan promises either 5 times or 20 times the usage of the standard $20-per-month Pro plan, but only inside rolling five-hour windows that are themselves subject to a weekly cap.

To find that out, a customer would need to click two separate hyperlinks buried in the purchase process, then visit a different page entirely to understand what the word "session" means. One Reddit user compared it to being handed a bigger fuel tank while the pump itself stays capped at one gallon every five hours.

"You've got to dive deep," said attorney Monica Vaca, who co-filed the complaint alongside Kati Daffan. Both spent careers at the Federal Trade Commission before entering private practice.

Who is being hurt, and how much are they paying?

The people most affected are heavy Claude users, often professionals who upgraded mid-project after hitting the cheaper plan's limits, then felt shortchanged after paying.

Vaca says she and Daffan heard repeatedly from people who felt they had to pay for top-tier AI subscriptions just to stay competitive at work, only to feel the product didn't deliver. "There is a long line of precedent on false advertising," she said. "You can't lie when you're marketing a product."

Anthropic pushed back in an earlier motion to dismiss, comparing the buried terms to flipping a product over to read the back label. The company said the session information was technically available to anyone willing to click through the purchase process.

Vaca called that framing unfair. Consumers can't audit what's inside the software. They have to trust what the marketing says up front.

What happens next?

The complaint was originally filed in July, then withdrawn and refiled as an expanded class-action, meaning it could eventually cover many subscribers, not just the original plaintiffs. Anthropic has not responded to the current filing.

Separately, Anthropic's latest model release, Claude 4.5 (sometimes called Fable 5.1 in internal references), acknowledged customer concerns about cost directly in its release notes, writing that the company was "addressing the feedback we've received from customers on price." That language suggests internal awareness that the pricing tension is real.

Takeaway: If you pay for any AI subscription, screenshot the usage terms on the day you sign up. If the product delivers less than those terms describe, that gap is exactly what consumer protection law exists to address.