Experts warn: nearly everyone in AI is susceptible to a honey trap scheme right now
Techies are building technology that can advance all aspects of the world — but can they protect their hearts? That's the vulnerability that honey pots exploit.
Experts are warning that nearly every individual involved in the artificial intelligence industry is at risk of being the target of a "honey trap" scheme. "Nobody briefs founders on counterintelligence," said Aaron Brown, a former CIA officer and founder of Lumbra. Traditionally, honey traps have been associated with politicians, military personnel, and national intelligence.
However, now many of the most valuable secrets are held by companies developing chips, models, and infrastructure. A spy seeks what a CEO can access, including roadmaps, export decisions, and customer lists.
OpenAI began warning its employees about potential honey trap attempts. Anyone with access to sensitive information, such as board members and executive assistants, can be a target. Honey traps often approach their targets as investors, customers, recruiters, or consultancies, gaining access through seemingly legitimate requests.
These schemes rely on fostering comfort in talking and can lead to blackmail once targets are compromised. AI has made honey trapping easier and cheaper, enabling the creation of fake social profiles, backgrounds, and even voice calls.
Signs of a honey trap include unusual patterns, persistent curiosity about work and colleagues, and requests for secrecy or boundary-pushing actions. Brown suggests recognizing when emotions may be influencing decisions and establishing clear boundaries between personal and professional life. Companies should promote open communication and discourage executive exceptionalism. Despite the industry's rapid growth, many tech founders are still vulnerable to the oldest human weakness: the desire to be understood by others.
Written by urgent.news from TechCrunch's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.