Anthropic did not sign a letter for open models that 200 firms signed

At the end of July, 2 large technology firms wrote a letter that tells all AI firms that some models must stay open. 24 companies agreed fast, and 200 more signed in the next week. The chief executive of Anthropic, Dario Amodei, did not agree to sign. 4 persons who know of it tell this.

The letter dispute

Anthropic and OpenAI told the US White House to prevent the release of some AI models, because the models are too dangerous. The 2 firms that wrote the letter saw good business in these models. OpenAI then signed the letter.

What rivals do

More than 20 persons in the industry, many with no name given, tell that rivals and investors:

• work as one group against the views of Amodei on very high AI risk.
• use their contact with US President Trump to stop regulators.
• speak in the open against how Anthropic does business.
• told one more executive of Anthropic that Amodei scares persons.

What is not known

The source does not give the view of Anthropic. It tells that Anthropic cut some customers and rivals off from its AI tools, but it gives no numbers. The effect on US policy is not known.

Insight: Anthropic incidents agree with both sides

Anthropic tells that some models are too dangerous to release to all. In October it told that its test agents used errors in government websites. A startup also went into its systems in less than 12 hours. This reads as a move to set the rules for open models. The incidents agree with its view, but rivals can use them against it. Leaders must prepare for rules that change.