AI company Anthropic is facing a class-action lawsuit. Users allege that the producer of Claude engaged in misleading marketing and false advertising with its subscription plans. The lawsuit was filed by two lawyers who previously worked at the Federal Trade Commission and is seen as the latest case of consumer advocacy amid the profit pressures faced by the AI industry.
The core dispute in the Anthropic lawsuit revolves around the Max subscription plan that Anthropic launched for Claude. The Max plan has two pricing tiers: US$100 per month (approximately HK$780) for "5x" the usage of the Pro plan, and US$200 per month (approximately HK$1,560) for "20x" the usage. However, the plaintiffs argue that these numbers are misleading—Anthropic's marketing page includes fine print stating that the actual promised multipliers only apply to five-hour time blocks and are also subject to weekly limits. When calculated, the total usage increase users actually receive is far lower than the advertised figures.
Attorney Monica Vaka, who is representing the plaintiffs, revealed in an interview that understanding the full meaning of these terms requires "digging deep." Users must click through at least two different links to find the definition of the term "session," and to understand what a "session" specifically refers to, they need to separately visit the Pro plan page for detailed explanations. The lawsuit cites a Reddit user's complaint: "The weekly cap is what the pricing page wants you to buy, but what actually determines whether you can continue using the service is that rolling five-hour window. It's like giving you a bigger gas tank but the gas pump only lets you add one gallon every five hours."
Anthropic's Max Subscription Plan Sparks User Discontent
The lawsuit also points out that many users, when encountering usage limits in their actual work, would immediately upgrade their plans, only to find out after payment that the actual usage was far less than expected. Vaka stated that such consumers often believed they would get a "significantly expanded" usage experience, but were disappointed after payment. Notably, Anthropic did not immediately implement the aforementioned limits when it launched the Max plan in April 2025, but only added these terms in August of the same year, a time when the company was intensifying its commercialization efforts to compete with OpenAI.
In the latest release of the Fable 5.1 model, Anthropic mentioned "responding to customer feedback on pricing," which indirectly confirmed user dissatisfaction with the pricing. Previously, in a motion to dismiss the lawsuit, Anthropic argued that the plan's limitation terms were provided via hyperlinks in the purchasing process, and consumers could access them if they were willing to click, "which is no different from reading the label on the back of a product." However, Vaka disagrees with this, stating: "Consumers have a hard time knowing what's inside that black box; they can only rely on the promises made by the marketers."
"They have no way to verify, no way to know exactly what they will get... they are essentially taking a risk by believing in a so-called 'honest market'."
Vaka and another plaintiff attorney, Katie Davan, have a combined 38 years of experience working at the Federal Trade Commission. She emphasized that false advertising has well-established legal precedents in the United States: "This is commercial speech; you cannot lie when marketing a product." She also mentioned that another important reason the two took on this case was that they had encountered many users who said they were forced to pay high subscription fees, fearing they would be left behind in the workplace, but ultimately felt that the service they received was not worth the price. The lawsuit was first filed in July of this year and was later withdrawn and resubmitted as a more complete class-action lawsuit.
Anthropic has not yet responded to the matter.