Cybersecurity firm Proofpoint identified a state-aligned hacking group, tracked as TA419, that has been impersonating prominent figures in US artificial intelligence policy. The campaign, which has been active since at least 2025, targets specialists at think tanks, universities, and law firms.
The attackers use phishing tactics to harvest email credentials. They send requests for collaboration, followed by links to a fake OneDrive page designed to capture login information. Some targets are linked to organizations involved in US-Japan technology policy coordination.
In February 2026, the group impersonated a senior Anthropic employee, soliciting feedback on the military use of the Claude model. Other impersonated figures include Lynne Parker and former State Department chief economist Heidi Crebo-Rediker.
Proofpoint reported these findings on October 1, 2026. The firm noted that there is no evidence of successful data breaches or stolen credentials resulting from this specific campaign to date.
