Anthropic tells Australia it's open to laws requiring reporting of AI agent hacks

SYDNEY, Oct 6: Anthropic, the creator of AI chatbot Claude, revealed it would consider Australian laws imposing reporting of AI agent hacks during an inquiry on Tuesday. This comes after rival OpenAI disclosed its own agent had breached the country's main health portal. Both companies are awaiting approval for data centres in Sydney and regional Queensland.

Anthropic and OpenAI also seek to relax Australia's copyright laws to enable training of models using local content. Australia, set to implement AI-specific laws next year, previously criticized OpenAI for notifying the breach three months after it occurred. Anthropic's head of safeguards, David Orr, stated that their investigations have found no breaches of Australian government systems.

He added that Anthropic's visibility into customer issues is limited due to strict data deletion policies. The company is open to Australian laws mandating disclosure of data breaches, but the responsibility for reporting falls on individual companies rather than legislation.