Anthropic sued for lying about Claude limits

4

Anthropic is getting hit with a lawsuit. It’s about the money you pay for Claude Max. The claim is simple: they are advertising limits they don’t actually enforce. Or rather, limits that don’t match reality.

Claude is an AI helper. You use it to code. To write. To think. The new suit says the two top tiers, Max 5x and Max a whopping 20x, cap you far before the math suggests. They call these caps usage limits. A “conversation budget.” The plaintiff says that budget is a lie.

Let’s look at the numbers.

The Pro plan costs roughly $20. Cheap. The Max 20x costs $200 a month. Twenty times the price. You expect twenty times the usage. You get maybe six to eight times. The Max 5x costs $100. Promise is five times. Delivery? Three and a half times.

It doesn’t add up.

Karl Khan filed the case. He’s the lead. He started using Claude for fun then got serious. Coding. Heavy lifting. By April 2026, he had upgraded to Max 20x. He burned through nearly 20% of a weekly budget in just five hours. He noticed the gap between the ad copy and the server limits.

The website is a black box, with no meaningful description of how usage is even calculated.

That’s from the complaint. Anthropic keeps the rules vague. No clear count on how many prompts you get per week. No definition of a “session.” Just advice on how to manage your stress while hitting a wall. When you hit the limit, you pay for credits. They reserve the right to cut you off at their discretion. A nice power move.

Frustrated users have taken to Reddit. One thread exploded with similar tales. “Extremely misleading” is a common phrase there. A bait and switch. The suit argues this isn’t just one person. It’s a pattern.

Context matters, I guess.

AI subscriptions are getting expensive. Nobody knows what they are buying. Transparency is thin on the ground. Yet people still buy it. Programmers do. “Vibe coders” do. The tools stick. Even while the political winds shift.

The Trump administration banned foreign governments from using Anthropic’s big tools. Shut some models down temporarily. Compliance issues. Anthropic keeps pushing updates anyway. They want to IPO. Space X just did the biggest IPO in history. OpenAI is next. Everyone is rushing to the cash register.

A rep didn’t respond. Probably busy.

The class includes anyone who bought Max 5x or 20x since April 2025 That’s a wide net. The complaint puts the stakes high. Over $5 million. Maybe more. We don’t know the final tab yet. But the anger is there. The discrepancy is there.

Whether this changes anything remains to be seen. Anthropic moves fast. Lawsuits move slow.

Попередня статтяApple Rips Out Old Search. Builds New One.
Наступна статтяToday’s Connections Trapdoor: Hidden Ships and Bad Cereal