Key facts
- The Trump administration ordered Anthropic to block foreign nationals from accessing its advanced AI models.
- Anthropic complied by suspending all customer access to its Fable 5 and Mythos 5 models.
- The order cited national security concerns and potential 'jailbreak' risks.
- Amazon, an investor in Anthropic, reportedly raised initial security concerns.
- US and European officials are now discussing AI model access following the dispute.
The Trump administration has ordered AI firm Anthropic to block foreign nationals from accessing its latest advanced models, citing national security concerns and the potential for 'jailbreaks' that could bypass safety guardrails. Anthropic complied by suspending all customer access to its Fable 5 and Mythos 5 models, which were recently released.
The dispute arose after Amazon, a significant investor in Anthropic, reportedly raised concerns about the models' security. While Anthropic argued that the identified security flaws were narrow and that a complete model recall was unwarranted, the administration issued an export control order.
This action has prompted discussions between US and European officials regarding access to AI models and the establishment of guidelines. The situation marks the latest in a series of disagreements between Anthropic, which emphasizes AI safety, and the Trump administration, including prior disputes with the Pentagon over contract terms.