The fallout from the allegations against Nvidia CEO Jensen Huang has exposed a blind spot across Silicon Valley. Security experts say the people most at risk from honey traps are no longer just politicians, soldiers and intelligence officers. They are founders, engineers and executives at AI companies, and many of them have never been trained to spot the threat.

“Nobody briefs founders on counterintelligence,” said Aaron Brown, a former senior CIA officer and current founder of the software company Lumbra, in an interview with TechCrunch.

What are honey traps and who is at risk?

Honey traps have long been associated with government workers who hold security clearances. Those people go through training that teaches them to expect an approach from a spy or another attacker. Private companies rarely offer anything similar. Brown argues that the most valuable secrets now sit inside companies building chips, AI models and infrastructure, which makes the private sector a natural target.

A spy, Brown said, “wants what the CEO can touch. The roadmap, the export decisions, the customer list, who gets the technology. The CEO is the shortest path to all of it.”

The allegations against Huang center on a woman named Alice Huang. She is a former executive of a Singapore neocloud company who has been accused of being a Chinese spy. She has also been accused of smuggling billions of dollars’ worth of restricted Nvidia chips into China. Huang has been accused of falling for a honey trap through an affair with her. The reporting describes those claims as allegations, and they have not been proven.

Jokes spread across the internet after the story broke, with users speculating about which tech chief executive might be next. The experts say the concern goes much further than the top job. The Wall Street Journal reported that OpenAI began warning its employees about outsiders who try to get close to them in search of company secrets. Anyone with access to sensitive information can become a target, from board members to executive assistants.

The popular image of a honey trap involves romance, but Brown said that is often the wrong picture. Attackers typically are not lovers. Friendship, he said, is the oldest honey-trap scheme there is.

These approaches often arrive looking like ordinary business. An attacker may pose as an investor, a customer, a recruiter or a consultant. The requests can appear reasonable on paper, and the target does not need to fall in love for the plan to work. “They need you to feel comfortable talking,” Brown said. “They last longer because nobody’s looking.”

Carpenter pointed to an irony in the tech world. Founders are building tools to automate and digitize much of daily life, yet they remain exposed to one of the oldest weaknesses there is. Attackers exploit the human need to be understood by others.

“Why fight through a firewall when you can get close to the person who can open doors, influence decisions, or make requests nobody wants to question?” Carpenter said.

Once a target realizes they have been compromised, the danger can grow. Attackers may then use what they have learned to blackmail their victims.

Can AI make honey traps easier to run?

AI has also lowered the barrier to entry. Planning a scheme once required months or years of research, careful construction of a fake persona, detailed knowledge of the target and a carefully executed plan. Today, Carpenter said, attackers can use AI to research a target, generate fake social media profiles and fake backgrounds, and even produce fake voice calls.

“AI can help sustain an entire fictional backstory,” Carpenter said. “The technology helps set the stage, but the deception ultimately plays out through human relationships.”

Brown said the AI industry “became a national security industry faster than its culture caught up.” He also warned that the danger is hard to see from the inside. “Part of being honey-trapped is not knowing,” he said. “It doesn’t feel like manipulation. It feels genuine, even lucky.”

How can executives spot a honey trap?

Carpenter recommended starting with unusual patterns. He pointed to “coincidental encounters that happen a little too often.” Other warning signs are easier to spot, such as persistent curiosity about work and colleagues, requests for secrecy about meetings, and pressure to make introductions or share information. A bigger red flag is a series of requests that keep growing and push past what a person would normally be comfortable doing.

Brown offered a simple test. “Would this person still be around if you lost your job tomorrow?” he asked. Carpenter suggested that people also consider when their emotions might be shaping a decision. He recommended drawing a firm line between personal and professional lives.

Company culture matters as well, the experts said. Employees need to feel free to raise concerns, challenge unusual requests and escalate possible threats. Carpenter said companies should work to “eliminate executive exceptionalism,” the idea that rules do not apply to people in power. “One of the most dangerous security cultures is one where nobody feels comfortable questioning the CEO,” he said.

Brown said leaders should not respond to the threat by shutting down their ambitions. “Founders need to take risks, and they need to explore business opportunities,” he said. “Paranoia doesn’t help. Recognizing you’re a target does.”

Questo articolo è stato scritto con l'assistenza dell'IA.
News Factory APP - notizie agentiche per potenziare il tuo SEO e AEO.