
KEY POINTS
- Experts told TechCrunch that AI founders, executives and staff are exposed to honey-trap schemes aimed at obtaining company secrets.
- AI tools now make it cheaper to research targets and to create fake profiles, backstories and voice calls.
- Specialists advise watching for unusual patterns, separating personal and professional life, and letting staff question executives.
Security specialists are warning that people working in artificial intelligence face a growing risk of honey-trap schemes, in which an attacker builds a close relationship with someone to obtain information or leverage, TechCrunch reported on November 20, 2026.
The warning follows allegations involving Nvidia chief executive Jensen Huang, which TechCrunch describes as a honey-trap scandal that is still unfolding. Those claims are allegations, and the outlet uses them mainly as context for a broader argument about how exposed the industry has become.
Aaron Brown, a former senior CIA officer who now runs the software company Lumbra, told TechCrunch that nobody briefs founders on counterintelligence. He explained that honey traps have traditionally been tied to politicians, military personnel and intelligence staff, who hold clearances and are trained to expect an approach. Now, he said, many valuable secrets sit inside companies that build chips, models and infrastructure. A spy, in his words, wants what the CEO can touch, such as product roadmaps, export decisions and customer lists.
The risk is not limited to chief executives. TechCrunch cited a Wall Street Journal report that OpenAI has begun warning its employees about outsiders who try to get close to them in search of company information. Anyone with access to sensitive material, from board members to executive assistants, could be a target.
Brown said the approach is often not romantic. Friendship, he noted, is the oldest honey-trap method. Attackers may pose as investors, customers, recruiters or consultancies, with requests that look legitimate. They only need the target to feel comfortable talking. Once someone realizes they have been compromised, the attacker may be able to use that as a means of blackmail.
AI is making these schemes easier and cheaper, according to the report. Preparing one once took months or years of research and persona building. Now AI can help research a target and generate fake social profiles, invented backgrounds and even fake voice calls. Perry Carpenter, a criminologist and chief deception strategist at KnowBe4, said AI can help sustain an entire fictional backstory, but the deception still plays out through human relationships.
The experts listed several warning signs: coincidental encounters that happen too often, persistent curiosity about work and colleagues, requests for secrecy about meetings, pressure for introductions or information, and requests that keep growing. Brown suggested a simple test: would this person still be around if you lost your job tomorrow?
On the organizational side, Carpenter recommended a clear boundary between personal and professional life and an end to what he called executive exceptionalism, the idea that rules do not apply to people in power. He also warned that a culture where nobody feels able to question the CEO is among the most dangerous for security. Brown added that paranoia does not help founders, but recognizing that they are targets does.