Dale Caldwell, New Jersey's former lieutenant governor, has turned to artificial intelligence to dispute the findings of a sexual harassment investigation that preceded his resignation. Caldwell maintains he is the victim of unfair targeting and says multiple AI agents support his position. He made the case during an appearance with host Rob Nelson, describing how he fed the investigation's report into several AI platforms.
Speaking with Nelson, Caldwell said he ran the report through AI and that the tools were asked what their findings would be. According to Caldwell, the AI systems raised the question 59 times, and none of them indicated there would be a finding of sexual harassment. He described the process as having "AIed" the report using multiple platforms.
The episode highlights a growing practice of individuals submitting official documents to chatbots for a second opinion. AI chatbots are widely known to be sycophantic, a tendency that leads them to tell users what they want to hear rather than deliver neutral assessments. That dynamic makes the tools a questionable arbiter for contested findings.
In Caldwell's case, the incentive cuts against any suggestion he was seeking a sympathetic answer. It appears highly unlikely that he would have wanted an AI to confirm that he engaged in sexual harassment. The platforms instead returned the conclusion he has publicly advanced since leaving office.
Caldwell resigned over the sexual harassment investigation but has continued to insist on his innocence. His account of the AI review now forms part of that defense. The former lieutenant governor has not backed away from his claim that he is being unfairly targeted.
The story sits at the intersection of two trends that are drawing attention in the United States: the use of AI tools in political and legal disputes, and persistent concerns about the reliability of chatbots when users put leading questions to them. For a US audience watching how AI is applied to public accountability matters, Caldwell's approach offers an example of the technology being enlisted to challenge an official investigative record.
The Verge published its report on the matter on October 4, 2026.
More AI news from TechManNews.
