
Cyberattacks targeting technology companies, government agencies and educational institutions are not new. However, a recent campaign highlights an emerging area of concern: threat actors are specifically targeting professionals involved in Artificial Intelligence (AI) policy, research, development and strategic planning.
According to research published by Proofpoint, a threat actor identified as TA419, and reportedly associated with Chinese intelligence interests, conducted a credential-phishing campaign targeting AI experts across multiple sectors. The campaign was observed from April 2025 through July 2026, with victims reportedly including U.S. and Japan-based government organizations, universities, legal organizations, defense contractors and other critical-sector entities.
The campaign is particularly significant because it demonstrates how traditional phishing techniques are being adapted to target individuals with access to highly valuable AI-related information.
Exploiting trust through AiTM Phishing
One of the techniques associated with the campaign is Adversary-in-the-Middle (AiTM) phishing. Unlike conventional phishing, where victims are typically directed to a fake login page designed to steal usernames and passwords, AiTM attacks can place an attacker between the victim and the legitimate authentication service.
In a typical AiTM scenario, a victim receives a convincing email containing a link to what appears to be a legitimate website or authentication portal. When the victim attempts to log in, the attacker can proxy the authentication process and potentially capture credentials or session information, including authentication cookies. This can allow attackers to bypass some forms of authentication and gain access to cloud services or corporate accounts.
The social-engineering component of this campaign is equally important. TA419 reportedly impersonated former White House officials, policymakers and other influential individuals to establish credibility with potential victims. Rather than relying on a single malicious email, attackers engaged targets through a series of communications designed to make the interaction appear legitimate.
Why AI Experts are valuable targets
The motivation behind targeting AI specialists goes beyond simply obtaining access to an individual’s email account. Professionals working on AI policy, research and strategic planning may have access to sensitive information concerning AI governance, model development, security controls, technology roadmaps, supply-chain strategies and government policy discussions.
Attackers reportedly invited victims to participate in a fictitious AI Policy Advisory Committee or requested their input for discussions associated with the Senate Committee on Foreign Relations, particularly concerning AI development, controls and supply-chain monitoring.
If successful, such approaches could provide attackers with access to confidential communications, internal documents, research material and information used in the development or deployment of AI systems. Compromised accounts could also potentially serve as launching points for further attacks against colleagues, partner organizations or connected cloud environments.
A changing Cybersecurity Landscape
The campaign demonstrates that AI is becoming an intelligence target as well as a technology. Organizations adopting AI need to consider not only the security of AI models and infrastructure but also the people who design, govern and influence these systems.
Organizations should strengthen phishing-resistant authentication, session protection, email security, identity monitoring and user awareness programs. Security teams should also pay particular attention to unusual authentication activity, suspicious OAuth applications, abnormal mailbox access and unexpected requests involving sensitive AI research or policy information.
As AI becomes increasingly integrated into government, legal, academic and commercial operations, protecting the experts who develop and govern these technologies will become an increasingly vital part of the broader cybersecurity strategy.
Join our LinkedIn group Information Security Community!











