AI Deepfakes Are a Growing Threat to Financial Advisors
AI Deepfakes Are a Growing Threat to Financial Advisors
Publish Date: 2026-06-03 11:18:00
Source Domain: www.wealthmanagement.com
On a set of large screens on both sides of the stage set for a panel discussion at the BNY INSITE Conference, in Aurora, Colo., Paul Caulfield, the chief compliance officer for Insigneo Advisors, appeared and addressed the audience.
He began by saying he was not Paul Caulfield.
“I am a synthetic version of him. I’m here today to explain that all is indeed not what it seems. Consider this. Consider that I was a synthetic representation of your own high-net-worth client,” the face said. “Better yet, consider that I was a synthetic representation of an ultra-high-net-worth prospect, someone you do not know well enough. How would you react? Would this seem convincing?”
Onstage, the real Paul Caufield told audience members he’d made the video in about 15 minutes spread over a day, and he showed it as an example of how easily artificial intelligence can be used to defraud seasoned financial advisors.
Caulfield was joined by Jeff Pollack, the interim global lead for identity and access management at BNY, and Keith Lanton, the president of the New York-based firm Herold & Lantern Investments, who detailed his firm’s experience with a cyberattack earlier this year.
While Lanton said the firm was lucky that hackers were not after personally identifiable information and that they were not successful in their goal, he said he quickly “got religion” to make sure the firm’s defenses were more robust than before, as bad actors’ skill and technical tools were only bound to get better.
“If you think about the fog of war and what’s going on, we didn’t know that we shut them down when we shut them down. We didn’t know the damage had been limited,” Lanton said. “So I say to each of you in this room, think about what your policies and procedures are. Think about what you would do if suddenly you discovered that something was infiltrated that you didn’t know could be infiltrated—and what you would do about…
