Anthropic Hit With Class Action Lawsuit Over Claude Max Subscription Usage Limits

Summary

Anthropic is facing a proposed class action lawsuit in federal court over its Claude Max subscription tiers, with a paying user alleging that the advertised 5x and 20x usage multipliers are impossible to achieve due to hidden weekly caps that throttle heavy usage. Plaintiff Karl Kahn, who upgraded to the $200 Max 20x tier for intensive coding work, claims that a five-hour coding session consumed 15% of his weekly quota, making the promised 20x boost mathematically unattainable.

The lawsuit, filed in the US District Court for the Northern District of California, targets both the $100 Max 5x and $200 Max 20x tiers and seeks refunds for all subscribers since launch in April 2025, plus damages under consumer protection law. The core complaint centers on Anthropic's dual-cap system—a five-hour rolling window limit plus a separate weekly cap—which Kahn argues were not clearly disclosed to users, making the advertised multipliers effectively unachievable during sustained work.

Anthropenic introduced the controversial usage caps in August 2025 after power users ran Claude Code continuously, initially claiming the restriction would affect fewer than 5% of subscribers. The timing and implementation have drawn widespread frustration, highlighting a fundamental tension in AI pricing: the extremely high cost of LLM inference makes fixed-price "unlimited" tiers economically unsustainable, yet marketing such tiers with specific multipliers creates false expectations. Similar consumer suits typically settle with modest refunds and clearer disclaimers rather than removal of the caps themselves.

• The case reflects the industry's broader struggle between sustainable AI economics and transparent customer communication about service limitations

Editorial Opinion

This lawsuit exposes a critical problem in how AI companies market subscription tiers: Anthropic advertised specific usage multipliers (5x and 20x) without adequately disclosing that technical caps would render those multipliers effectively meaningless for sustained work. While the company's core constraint is legitimate—LLM inference is genuinely expensive and unlimited fixed-price tiers are economically unsustainable—the marketing approach prioritized conversion over clarity. The case is a wake-up call for the AI industry: customers will not accept a disconnect between marketing promises and actual capability, especially when paying premium prices.