Foreign Intelligence Operatives Impersonate U.S. Officials in Phishing Campaign Against AI Policy Researchers
Chinese hackers launched phishing campaigns impersonating prominent U.S. officials and private sector leaders to target artificial intelligence policy experts at universities, think tanks, and law firms. The attackers used fake invitations to policy advisory committees and legislative working groups to solicit responses before deploying credential-stealing links. Cybersecurity firm Proofpoint attributed the operation to a China-aligned hacking group focused on obtaining intelligence related to AI policy and export controls.
The impersonation campaign demonstrates a deliberate strategy to exploit professional credibility. By mimicking officials with genuine connections to AI policy circles—including someone from a leading AI company—the attackers established false legitimacy that made their requests appear routine. The timing and subject matter, including military applications of AI systems and export control mechanisms, suggest coordinated intelligence gathering aimed at understanding U.S. policy development processes rather than random credential harvesting.
TA419's operational pattern reflects sustained focus on sectors critical to national competitiveness. The group's documented interest in defense, energy, and foreign policy alongside AI expertise indicates an effort to map the landscape of American policy influence. The February targeting of an Anthropic employee signals particular interest in how private sector AI capabilities intersect with government applications and restrictions.
This campaign could undermine the effectiveness of policy development by creating friction between researchers and legitimate officials. Experts may grow hesitant to engage openly in policy discussions if impersonation becomes a persistent threat. The targeting of think tanks and universities may discourage public-private collaboration on sensitive issues, potentially slowing policy innovation during a period when international competition in AI development remains intense. Organizations across sectors may need to invest in authentication protocols that balance security with maintaining productive professional networks.