Anthropic Claude Max 计划遭扩大集体诉讼，被指误导重度用户

Anthropic表示高频用户是其业务的关键——即便这意味着要切断其它受欢迎的应用程序，如OpenClaw，它也优先考虑这些用户。但其中一些客户表示，Anthropic误导他们，让他们相信高端订阅计划会比实际获得的更多。

Anthropic为Claude推出的Max计划是20美元/月Pro计划的升级版。它有两个定价选项：每月100美元可获得Pro计划使用限制的“5倍”，或每月200美元可获得“20倍”。但诉讼称，这些数字在Anthropic的营销图中展示时，以细小字体误导性地表达。Anthropic实际上表示，你将获得比Pro计划更多5倍或20倍的使用量，但这只是保证在每次五小时的时间块中有效，并且受到每周限制——投诉称，这导致实际总使用量的增加远小于广告宣传的数字。

要理解这一点，“你必须深入研究，”Vaca说。这需要点击两个不同的超链接才能看到“session”一词，然后要了解“session”的含义，你还需要单独访问Pro计划网页以获取详细定义。

关于Max条款的投诉已经出现在网上。一位Reddit用户写道：https://www.reddit.com/r/ClaudeAI/comments/1w4sul5/max_5x_100month_current_session_100_used_weekly/，“每周配额是定价页面让你以为自己购买的内容。滚动的五小时窗口才是真正控制你是否可以工作的因素……就像给某人更大的油箱，但每五小时燃油泵只能供应一加仑。”

在过去八个月或更长时间里，随着盈利压力的增加，AI客户经常抱怨顶级实验室将其高昂的运营成本转嫁给客户。尽管Anthropic在2025年4月宣布了Max计划，但直到几个月后的8月才实施所谓的误导性每周限制，当时公司正在努力与OpenAI竞争。在Anthropic最新发布的模型Fable 5.1中，公司暗示了客户更广泛的成本关注，写道在新模型定价中“正在处理我们收到的客户关于价格的反馈。”

该投诉最初于七月提交，随后被撤回，并重新作为扩展的集体诉讼重新提出。在驳回早期案件的动议中，Anthropic辩称，它并不一定在其营销中遗漏会话限制；相反，它表示如果消费者知道去哪里查找，技术上是可以获取的。Anthropic在驳回动议中写道：“访问这些澄清信息……只需要点击购买过程中可用的超链接——这在数字上相当于翻过产品查看背面的标签。”

Anthropic 未对置评请求作出回应。

Vaca不同意这是那么容易理解条款。“这对消费者来说很难——他们不知道那个黑盒子里有什么，”她说。“他们必须依赖营销者提供的说明。没有办法让他们进行审计。也无法准确知道他们将获得什么……他们基本上是在冒险，并在相信一个诚实的市场。”

她补充说，案件对她“具有吸引力”的另一个原因是，她和Daffan开始听到一些人觉得，为了避免在就业市场上变得无关紧要，他们需要支付高价订阅AI服务。但她说，很多时候，这些人觉得自己并没有得到所支付的价值，这对每月花费100到200美元的服务来说令人沮丧。

“[Daffan]和我在联邦贸易委员会合计工作了38年，”Vaca说，并补充道，“关于虚假广告有一长串的先例。这是商业言论。你在营销产品时不能撒谎。”

即使Anthropic在其网站的某处有这些限制条款，Vaca表示，也不应该由消费者去主动寻找。

“这是一种‘买者自慎’的方式，”她说。“这对人们来说真的不公平。”

Anthropic says power users are key to its business — it’s prioritized them even when it means cutting off other popular applications, like OpenClaw. But some of these same customers say Anthropic misled them into believing they’d get more out of a top-tier pricing subscription than they did.

Anthropic’s Max plan for Claude is an upgrade to the $20/month Pro plan. It has two pricing options: $100 per month for “5x” the usage limits of Pro or $200 per month for “20x”. But the lawsuit alleges that these numbers, advertised in Anthropic’s marketing graphics, are misleadingly couched in fine print. Anthropic actually says you’ll get 20x or 5x the usage that the Pro plan offers, but it’s only promised for five-hour chunks of time that are subject to a weekly limit — which the complaint alleges adds up to a much smaller total increase in usage capacity.

To understand this, “you’ve got to dive deep,” Vaca said. It requires clicking two different hyperlinks to see the word “session,” and then to see what “session” meant, you’d need to separately go to the Pro plan webpage for an expanded definition.

Complaints about Max’s terms have cropped up online. One Reddit user wrote：https://www.reddit.com/r/ClaudeAI/comments/1w4sul5/max_5x_100month_current_session_100_used_weekly/, “The weekly allowance is what the pricing page makes you think you’re buying. The rolling 5-hour window is what actually controls whether you can work … It’s like giving someone a bigger gas tank while keeping the fuel pump limited to one gallon every five hours.”

Over the past eight months or more, as pressures to turn a profit increase, AI customers have frequently complained that top labs are passing on their steep operating costs. Although Anthropic announced the Max plan in April 2025, it didn’t impose the allegedly deceptive weekly limits until a few months later in August, as the company pushed to compete with OpenAI. In Anthropic’s latest model release, Fable 5.1, the company hinted at broader cost concerns among customers, writing that it was “addressing the feedback we’ve received from customers on price” in the new model’s pricing.

The complaint was filed in July before being withdrawn and refiled as an expanded class action. In a motion to dismiss the earlier case, Anthropic argued that it didn’t necessarily omit the session limits from its marketing; rather, it says it’s technically available to the consumer if they know where to look. “Accessing this clarifying information … required nothing more than clicking hyperlinks available in the purchase process—the digital equivalent of flipping a product over to read the back label,” Anthropic wrote in the motion to dismiss.

Anthropic did not respond to a request for comment.

Vaca disagrees that it’s that easy to understand the terms. “This is hard for consumers — they don’t know what’s in that black box,” she said. “They have to rely on the claims the marketer is giving. There’s no way for them to audit it. There’s no way for them to know exactly what it is they’re going to get… They’re basically taking a leap, and they’re believing in an honest marketplace.”

She added that another reason the case was “compelling” for her is that she and Daffan started hearing from people who felt they needed to pay high-priced subscription costs for AI services in order to avoid becoming irrelevant in the job market. But oftentimes, she said, they felt they weren’t getting what they paid for, which was frustrating for a service that costs between $100 and $200 per month.

“[Daffan] and I worked at the Federal Trade Commission for 38 years combined,” Vaca said, adding, “There is a long line of precedent on false advertising. It’s commercial speech. You can’t lie when you’re marketing a product.”

Even if Anthropic has the stipulations somewhere on its website, Vaca said, it shouldn’t be up to the consumer to have to seek that out.

“It’s a ‘buyer beware’ approach,” she said. “And that’s really not fair to people.”