While the world keeps talking about the power of Anthropic’s new AI model, Fable 5, some observers are urging consumers to pay attention to the company’s consumer terms, claiming the firm has the right to send lawyers after users and even make them pay for the legal process. The truth is a little more complicated, though.

“If you’re one of the numerous people posting about Anthropic’s dystopian ways and you’re thinking about getting Claude to help you write that post... don’t!” Arnaud Bertrand, a French entrepreneur, said on X on Wednesday.

Indeed, Anthropic’s consumer terms for Claude AI explicitly ban users from taking any action that exposes the company to “reputational harm,” giving the platform broad discretion to ban users or terminate accounts based on subjective interpretation of behavior.

Injunction as a weapon

Moreover, if consumers do expose the company to those harms, they had actually pre-agreed in the terms that they wouldn’t oppose Anthropic’s injunction, which it claims is the “best remedy for any such breach.” The company doesn’t even need to prove actual damages.

According to Bertrand, this means that Anthropic can simply go to a judge in a friendly jurisdiction and “make you pay for everything.”

That’s because under Section 11 of Anthropic’s Consumer Terms, users agree to indemnify Anthropic for “any and all liabilities, claims, damages, expenses (including reasonable attorneys’ fees and costs), and other losses arising out of or related to your breach or alleged breach of these Terms.”

“In other words, if you use Claude to help you talk shit about Anthropic publicly, their terms say you pay their lawyers to go after you, and you’ve already pre-agreed you’ve lost the case,” Bertrand claims.

Moreover, even if a judge throws out Anthropic’s case against you, the company’s maximum liability appears to be set at $100. They have also capped damages to the amount you paid for the service in the last six months.

Quite a few caveats apply

Sounds bad, doesn’t it? A pinch of salt is needed, though. Bertrand is not an impartial observer simply wishing to warn naive AI chatbot users about alleged risks.

Bertrand, heavily quoted on Chinese state media, claims that he’s busy every day debunking anti-China propaganda.

Under that same post about Anthropic’s consumer terms, he says: “Chinese open source models aren’t even remotely like this.”

That’s not true. Chinese AI models, such as those from Qwen and DeepSeek, also enforce heavily one-sided terms for consumers and businesses, forcing users to sign broad indemnity clauses that shift all legal and financial liability onto the consumer.

Besides, what Bertrand is claiming applies only to US jurisdictions. For instance, Anthropic’s terms for consumers in the United Kingdom are “governed by English law.”

Finally, Anthropic’s rivals like OpenAI and Google aren’t really any better. Their terms mirror Anthropic’s indemnity waivers, also disclaiming responsibility for indirect damages and limiting liability to roughly $100 or subscription amounts.

Considering legal action against users would, of course, be a massive PR disaster for any large AI firm. However, the clauses do exist, and the user should know exactly what they’re signing up for.

---

Unlock more exclusive Cybernews content on YouTube.