U.S. lifts export controls on Anthropic's Fable 5 and Mythos 5 models

The United States Department of Commerce has removed export controls that had been placed on artificial intelligence models developed by the U.S. company Anthropic. The decision affects the advanced AI systems known as Claude Fable 5 and Mythos 5.

These export restrictions were implemented in early June as part of a broader regulatory framework aimed at limiting access to sophisticated AI systems by entities deemed sensitive. Concerns centered on the potential for these powerful models to be misused in areas such as cybersecurity and other critical infrastructure. The lifting of these controls follows negotiations between Anthropic and the Commerce Department.

Anthropic announced that it will commence restoring access to these AI models for authorized users starting Wednesday. The company had previously indicated significant progress in discussions with the U.S. government, which had expressed fears that the models could be manipulated to facilitate cyberattacks. The reversal of the restrictions allows for a phased reopening of access.

In parallel, Anthropic is engaged in legal disputes with the Pentagon in the United States concerning restrictions on the use of its models in military contexts. These ongoing conflicts have involved preliminary judicial decisions regarding potential administrative retaliatory measures.