China-Aligned Espionage Group Targets US AI Policy Experts
A China-aligned espionage group has been identified as posing as a former White House official and a prominent economist to gain access to the cloud accounts of AI policy experts in the United States, according to findings from Proofpoint. This group, tracked as TA419, executed multiple credential phishing campaigns throughout July 2026.
Beginning on July 8, 2026, TA419 impersonated Lynne Edwards Parker, who previously served as the Principal Deputy Director of the White House Office of Science and Technology Policy, and later Heidi Crebo-Rediker, an economist and foreign policy expert. In a separate incident in February 2026, the group impersonated a senior employee from Anthropic to target an AI policy analyst at a US think tank. The email sent during this campaign had the subject line "Request for Feedback on Military Integration of Claude," referencing discussions surrounding the US military's use of Anthropic's Claude models.
The initial emails sent by TA419 did not contain any malicious links. Instead, they invited recipients to join a fictitious "AI Policy Advisory Committee" or to contribute to a Senate Committee on Foreign Relations report concerning AI export controls and supply chains. If a target responded, TA419 would follow up with a shortened URL that purported to provide additional information, which ultimately led to a fake OneDrive credential phishing page designed to capture the target's cloud account credentials.
The phishing link was structured to first pass through a domain controlled by the group, which employed a Cloudflare Turnstile check behind a fake OneDrive loading screen. Victims were then redirected to a second domain hosting the phishing page, which displayed a OneDrive folder listing with lure documents. Clicking on any document or the OneDrive prompt to sign in would open a fake Chrome login window, created using an open-source tool known as Frameless BitB.
Mechanics of the Phishing Attack Unveiled
Credentials entered into this window were relayed to Microsoft's servers, allowing the attackers to capture passwords, multi-factor authentication codes, and pass conditional access checks. A custom script within the phishing kit tracked the victim's progress, providing the attackers with a live view of the session, accepting the "Keep me signed in" prompt, and submitting one-time codes as soon as they were validated. The kit was designed to capture session cookies created at the end of the login process, granting the attackers access to the target's account.
Case Closed
TA419 consistently utilized Cloudflare's content delivery network to obscure the backend hosting IP addresses for its domains, which were typically registered through NameSilo. The group's credential phishing domains often mimicked file-sharing sites and cloud services. Occasionally, they registered domains that impersonated specific organizations, such as heritiages[.]org for the Heritage Foundation and tw-koryu[.]org for the Japan-Taiwan Exchange Association.
This espionage activity likely supports broader Chinese intelligence objectives aimed at understanding ongoing developments within the US AI policy and regulatory landscape. It occurs amid heightened strategic competition, allegations of model distillation, and export controls involving the US and China.
