Fake Singapore-Based Think Tank Used AI Profiles to Target Intelligence Community
An entity identifying itself as a Singapore-based think tank, the “Strategic Asia Pacific Peace and Security Forum” (SAPPSF), has been identified as a front for a sophisticated influence operation utilizing artificial intelligence-generated personas to target security analysts and members of the intelligence community. According to a report by the cybersecurity firm Mandiant, the network employed fake identities to gain access to sensitive professional circles and disseminate curated narratives.
## Operational Tactics and AI-Generated Personas
The operation relied on the creation of high-fidelity, AI-generated profile pictures to establish the credibility of its staff. These personas were used to interact with experts in international relations and defense policy across social media platforms, specifically targeting individuals with influence in the security sector.
Mandiant researchers noted that while the group claimed to be based in Singapore, there was no evidence of a physical office or legitimate institutional registration in the country. Instead, the network functioned as a digital-only entity designed to facilitate contact with legitimate professionals. By posing as a think tank, the actors attempted to solicit contributions or engage in discussions that could provide them with insights into regional security policies or allow them to plant specific viewpoints.
## Targets and Institutional Reach
The campaign specifically sought out analysts, academics, and government-affiliated personnel who focus on the Asia-Pacific region. By masquerading as a research organization, the actors bypassed standard social media suspicion, as the targets were conditioned to engage with think-tank representatives as part of their professional duties.
The use of AI-generated imagery—often characterized by subtle, tell-tale glitches or inconsistencies in facial features—served as the primary vehicle for maintaining the appearance of a large, diverse team of experts. This approach allowed the group to sustain multiple, simultaneous conversations with different targets, creating an illusion of broad institutional reach.
## Detection and Attribution
The identification of the SAPPSF network marks a shift in how state-aligned actors are leveraging generative AI to conduct intelligence-gathering operations. Analysts from Mandiant highlighted that the campaign was not designed for mass-market disinformation but rather for “spear-phishing” at the intellectual level, where the goal was to build long-term rapport with individuals who possess specialized knowledge.
The investigation into the network remains ongoing as security firms continue to monitor for similar clusters of activity. No specific state actor has been definitively named as the architect of the SAPPSF campaign, and the group has maintained a policy of silence regarding its operational funding and physical headquarters. The relevant cybersecurity authorities have not yet announced official regulatory action against the platform, and the network’s digital presence continues to be scrutinized by international intelligence agencies.