Trump administration: now requiring AI companies to report and fix security breaches after Anthropic’s 'fraudulent' use of government systems - Axios