Anthropic Sued Over Claude Max's Hidden Usage Limits

A proposed class action says Anthropic's Claude Max plans deliver far less usage than advertised, with one user burning 15% of a weekly cap in one session.

4 min readEAEvgenii ArsentevEvgenii Arsentev · PhD

Anthropic is facing a proposed class-action lawsuit claiming its Claude Max subscriptions deliver far less usage than advertised. Karl Kahn, a Washington, DC resident, filed the suit in federal court on June 15, alleging that 'the actual usage provided by the Max 5x and Max 20x plans is far below the advertised amount of usage.'

Anthropic markets the two Max tiers — $100 and $200 a month — as offering five and twenty times the usage caps of the cheaper Pro plan. Kahn says that promise did not hold up: after upgrading to Max 20x in April 2026, he reported burning through 15% of his weekly allowance in a single five-hour session, then hitting weekly limits soon after. The suit seeks class-action status for US customers who bought Max subscriptions since April 2025. Anthropic declined to comment.

Why a rate-limit lawsuit matters

This is the friction point of the whole AI subscription model laid bare. Running a frontier model is expensive per query, but consumers are sold flat monthly plans with reassuring multipliers like '20x.' What '20x' actually buys is rarely spelled out in tokens or messages, so two people on the same plan can have wildly different experiences depending on how heavy their prompts are. If you have ever been coding along and suddenly been told you are rate-limited for days, you have felt exactly what this lawsuit is about.

I lean on Claude Code heavily, so I have watched these caps move and tighten over the past year — and the honest problem is that 'usage' is defined in a unit you cannot see while you work. Whether or not Kahn wins, the suit pushes on something real: subscriptions priced as 'unlimited-ish' but metered invisibly. The likely outcome across the industry is clearer disclosure — actual token budgets and live usage meters — rather than vague multipliers that mean different things to different people.

What I'd actually do

Before paying for any AI plan, find the real limits — search the provider's docs for token or message caps, not just the marketing multiplier. Watch your usage in the first week and note when you hit a wall. If your work depends on not getting cut off mid-task, keep a second provider or a local model ready, and never assume 'Max' means 'unlimited.'

#ai#anthropic#claude#subscriptions#legal

Related guides

EAEvgenii Arsentev

Author

Evgenii Arsentev

PhD · Chief Product Officer at a healthtech company

Want to actually build this?

Guides explain. The free course transforms — personalized, gamified, and built to get you shipping fast.

◉ Start the free course

Source: engadget.com