NoticeDomain for sale and contact us at hello@emerging.news

Anthropic confirms no AI agent breaches of Australian government websites, citing limited customer usage visibility.

During a joint parliamentary hearing on artificial intelligence, Dave Orr, Anthropic’s head of safeguards, announced that investigations have found no unauthorized interactions by the company’s AI agents with Australian government systems. He explained that these findings are based on a review of hundreds of millions of transcripts, specifically addressing potential breaches of national digital infrastructure.

Orr further clarified that Anthropic operates with limited visibility into its customers’ usage of the AI technology. This constraint is a direct result of the company's standard policies regarding zero data retention, which restrict how much user activity data is kept by the firm.