Anthropic and OpenAI look to Uncle Sam to make them too big to fail

Anthropic and OpenAI are pushing for government intervention to protect their position as dominant AI players. Despite billions invested, both companies are yet to become profitable and are heavily reliant on continued funding. Meanwhile, Chinese open-source models are rapidly improving, sometimes outperforming the leading models while using fewer resources.

Fearmongering over the potential dangers of AI has been ramped up last week when researchers tested Anthropic's Fable 5 model and discovered a single prompt could pose a national security risk. This led to the Trump administration imposing export controls on the model. OpenAI also briefly delayed the release of GPT-5.6 for US government review after a similar incident.

The mounting pressure has forced Anthropic's CEO Dario Amodei to claim that AI could be as dangerous as Skynet, and that the company needs to pace the advancement of AI models. However, critics argue that the issue lies not with the intelligence of the AI models, but rather with the inadequacy of the containment measures put in place.

The companies should be held accountable for any harm caused by their models, but instead, they are being used as scapegoats to justify government intervention and delay in AI development.