Anthropic, the $965 billion AI company, may have talked itself into trouble. According to a Financial Times analysis, the company warned about the dangers of advanced AI far more often than rival OpenAI this year. Now, critics say those warnings helped trigger a US ban on foreign access to its newest models.
The analysis found that five in every 1,000 words used by Anthropic in 2026 related to risk, regulation, or restrictions. The equivalent figure for OpenAI and its chief Sam Altman was eight times lower, at just 0.6 words per 1,000. The research looked at official statements, social media posts, and articles written by the company or its chief, Dario Amodei.
How warnings led to an export ban
Washington last week barred foreign nationals from using Anthropic’s latest models, Mythos and Fable. Some technologists have blamed the decision on the company’s repeated warnings about AI risks. According to FT analysis, the comparison has become politically charged after the ban was imposed.
The move came after Anthropic refused to remove safeguards that prevent its AI from being used in fully autonomous weapons systems or domestic surveillance, according to a Google Featured Snippet summarizing the situation.
What the ban means for Anthropic
The export ban means foreign nationals can no longer access Anthropic’s most advanced AI models. This is a significant restriction for a company that has positioned itself as a safety-first alternative to OpenAI. By warning so loudly about risks, Anthropic may have given regulators the very reason they needed to lock down its technology.
Critics argue that Anthropic’s strategy backfired. Instead of building trust, its constant focus on dangers may have convinced Washington that the models were too risky to share internationally. The ban now affects researchers, developers, and businesses outside the US who rely on Anthropic’s technology.
Our Take: A cautionary tale for AI companies
In our view, Anthropic’s situation is a clear warning for the entire AI industry. The company tried to be responsible by highlighting risks, but that honesty may have cost it access to global markets. This does not mean companies should hide dangers — but it does show that regulators listen closely to what AI firms say about their own products.
The real problem is that the US government used Anthropic’s warnings as a reason to restrict access, rather than as a reason to work with allies on shared safety standards. To put it plainly, Anthropic’s transparency was punished, not rewarded. Other AI companies should take note: talking about risks can have real-world consequences, especially when governments are looking for reasons to control advanced technology.
For now, Anthropic is left with a difficult choice — either tone down its warnings to regain global access, or stick to its principles and accept a smaller, US-only market.
