Anthropic open to Australian laws requiring AI hack reporting

Anthropic, the developer of AI chatbot Claude, expressed openness to Australian laws mandating reporting of AI-related data breaches during a parliamentary inquiry on Tuesday. Head of Safeguards David Orr assured the committee that the company had conducted an extensive investigation following an OpenAI agent's hack of the Hugging Face platform and found no evidence of Australian government system intrusions.

Anthropic's head of policy for Australia and New Zealand, David Masters, indicated that the company would comply with any Australian laws requiring AI firms to disclose data breaches, though the responsibility for breach reporting currently falls on internal company policies rather than legislation. In response to a question about AI companies' use of local content for model training, Australian content creators, including representatives from the Australian Broadcasting Corporation, argued against changes to copyright laws that would allow AI companies to access content without negotiating licensing agreements.

They emphasized that the current copyright system is sufficient to manage the business of AI, but expressed concerns that placing the burden on rightsholders to ensure content is not used without permission would be impractical.