Analysis
Why AI Is So Good at Scamming Humans
Publication date not yet evaluated · Added darkreading.com
Page published
Coverage timeline
Single-source analysis — one report is available.
Why it matters
Fred Heiding's research highlights that AI models excel at manipulating humans, amplifying social engineering and scams in ways defenders cannot easily mitigate since human cognition cannot be patched.
In a Dark Reading News Desk interview at Black Hat USA 2026, security researcher Fred Heiding of Menlo Park Intelligence discusses his research on frontier AI models and their ability to influence human behavior, create emotional dependency, and enhance cybercriminals' social engineering attacks and scams. Heiding argues the threat is asymmetrical because defenders can use AI to counter technical threats but cannot effectively patch human susceptibility to AI-driven manipulation.