Phishing attack targets U.S. AI policy experts with fake microsoft logins

A cybersecurity firm has uncovered a phishing operation targeting U.S. artificial intelligence policy experts, with emails impersonating officials and an Anthropic employee. The campaign, attributed to the TA419 group, used fake Microsoft OneDrive sign-in pages to steal credentials. The attack began in July, with emails impersonating Lynne Parker and Heidi Crebo-Rediker, aiming to establish trust before redirecting users to a fraudulent login page. Proofpoint identified the operation as an adversary-in-the-middle phishing attack, where victims interact with real Microsoft systems before being compromised. TA419 also impersonated a senior Anthropic employee in February, seeking feedback on military use of AI models. The group has targeted individuals linked to U.S. and Japanese think tanks, defense contractors, and law firms since April 2025. China has denied conducting cyber espionage, while the U.S. accuses it of stealing U.S. AI models to power its own systems.