Australian authorities and companies seek access to “dangerous” AI model

A new generation of AI models deemed “too dangerous” to release publicly is driving Australia’s efforts to encourage leading AI companies to establish a larger presence in the country.

AI company Anthropic’s chief counsel, Jeffrey Bleich, flew to Australia this week for meetings with senior government officials.

On the meeting agenda is access to Anthropic’s newest model, Claude Mythos.

Mythos and a similar new model from rival OpenAI are touted as being so capable of finding cybersecurity vulnerabilities that companies fear major disruption if the technology falls into the wrong hands.

Now Australian agencies and companies are pushing for access to the models, increasing the influence of the buzzy AI company as it considers where it will set up key hubs outside the US.

A federal government briefing obtained under freedom of information laws says Australia wants “major investment in Australia” and is willing to discuss a review of copyright laws as part of an agreement.

Anthropocene…