Anthropic 在诉讼中被指控对其每月 200 美元的“20 倍”计划撒谎
Anthropic Accused In Lawsuit Of Lying About $200 Per Month '20x' Plan

原始链接: https://www.zerohedge.com/ai/anthropic-accused-lawsuit-lying-about-200-month-20x-plan

针对人工智能公司 Anthropic 的一项联邦集体诉讼已被提起。诉讼指控该公司在高端“Max 5x”和“Max 20x”订阅套餐的使用限额方面误导了消费者。该诉讼由订阅用户卡尔·卡恩(Karl Kahn)在加利福尼亚北区联邦地区法院代表提起,诉状称,这些每月费用分别为 100 美元和 200 美元的计划,并未能实现比标准专业版订阅多出 5 倍和 20 倍的容量提升。 原告称,尽管他为了高强度的编码工作升级到了“Max 20x”套餐,但仍频繁遇到使用限制,被迫限制提问次数或购买额外额度,这与 Anthropic 的营销宣传相矛盾。该诉讼旨在为受影响的用户寻求退款,并要求法院裁定该公司的广告宣传存在欺诈行为。 此次法律行动发生之际,业界对人工智能不可预测的速率限制和计算资源稀缺问题普遍感到不满。由于安全担忧,联邦政府近期出台限制措施,禁止外国用户访问 Anthropic 最强大的“Mythos 级”模型,使局势变得更加复杂。Anthropic 目前尚未就该诉讼发表正式评论。

相关文章

原文

A federal class-action lawsuit filed Monday accuses Anthropic of misleading customers about the real usage limits on its high-end Claude AI subscriptions. The suit, brought on behalf of Washington D.C. subscriber Karl Kahn and others who bought the Max 5x and Max 20x plans since April 2025, claims the company oversold how much computing power buyers would actually receive.

The lawsuit - filed Monday in the Northern District of California on behalf of Washington DC resident Karl Kahn and others who subscribed to the plans since April 2025 - targets Anthropic's Max 5x and Max 20x tiers priced at $100 and $200 per month respectively. It accuses the company of misleading customers by advertising these plans as providing five and twenty times the usage capacity of the standard Pro subscription, when in reality the actual limits fall well short of those claims. The allegations draw heavily from emails Anthropic sent to subscribers in July 2025 that outlined the expected weekly usage allowances for each tier at the time.

According to the complaint, Kahn upgraded to the Max 20x plan in April of this year after increasing his reliance on Claude for coding work. He soon discovered he was exhausting his weekly limits rapidly, including burning through 15 percent of his allowance during a single five-hour session. The suit seeks refunds for affected customers and a judicial finding that Anthropic's marketing of the high-tier plans was fraudulent.

Kahn initially used Claude for personal tasks but later relied on it heavily for coding. After upgrading, he repeatedly hit usage walls and had to stop work, ration prompts, or buy extra credits to finish projects, according to the complaint. The lawsuit says the actual limits are difficult to predict and consistently lower than what was promised when the plans were marketed as giving five or twenty times the capacity of the standard Pro subscription.

"The actual usage provided by the Max 5x and Max 20x plans is far below the advertised amount of usage," reads the lawsuit, that claims Kahn "found himself needing either to halt his work, ration his usage, or purchase additional usage to ensure that he could complete his work." 

Anthropic has not commented on the suit, according to the Wall Street Journal. The company offers free access plus paid tiers, with the Pro plan running $17 to $20 a month. The higher Max plans were positioned for power users needing substantially more compute.

This lawsuit arrives amid mounting frustration with AI subscriptions and tokenomics. Power users and even large enterprises have complained for months about unpredictable rate limits, especially on coding workflows - with several documented cases of extreme overspending, including one unnamed Anthropic client (Amazon?) that racked up roughly $500 million in Claude charges in a single month after failing to cap employee usage.

Compute scarcity remains a core issue across the sector. A surge in demand earlier this year strained systems at Anthropic and rivals, producing outages and tighter limits even for paying customers. At the same time, companies are racing to launch new models ahead of expected IPOs while navigating new government restrictions. Days before this suit, the Trump administration banned foreign governments, companies, and individuals from accessing Anthropic's most powerful models after Amazon discovered a way to jailbreak the company's Fable AI into its unrestricted form - Mythos, forcing the company to shut off certain access to comply.

On Sunday, Anthropic execs scrambled to DC to triage the situation

Fable 5 launched on June 9 as the first broadly available "Mythos-class" model, the public-facing version of a system Anthropic had previously kept behind a vetted-access wall because of its cyber and biological capabilities. Mythos 5, the same underlying model with some safeguards removed, stayed reserved for cleared cybersecurity partners. Fable 5 was the middle path: Mythos-grade capability, Anthropic said, with guardrails strong enough for general release. The company put it on the API, made it generally available on Amazon Bedrock and GitHub Copilot, and folded it into Pro, Max, Team, and Enterprise plans at no extra charge through June 22.

联系我们 contact @ memedata.com