New Jersey Lieutenant Governor Uses AI to Contest Harassment Probe

Dale Caldwell, the former Lieutenant Governor of New Jersey, is attempting to dispute official findings that led to his resignation on September 25th. Following an investigation that concluded he committed sexual harassment and breached ethics standards, Caldwell has turned to generative artificial intelligence to validate his claims of innocence. The Scale Report understands that the former official presented this unconventional evidence during an interview with Rob Nelson on NJ PBS.

The Role of AI in Legal Defense

Caldwell reported that he processed the official investigation report through multiple AI platforms to test for bias. During the broadcast, he asserted that he ran the documents through these systems fifty-nine times and claimed none of the models produced a finding of sexual harassment. By leveraging these automated tools, he argued he was being unfairly targeted by the investigative process.

Reliability of Automated Analysis

Technology experts often caution that large language models are prone to sycophantic behavior, frequently mirroring the biases or desired outcomes of their human operators. Because these systems are designed to provide conversational responses rather than perform rigorous forensic analysis, they are not qualified to adjudicate ethics investigations or provide legal exoneration.

This incident highlights a growing trend where public figures attempt to use AI output as a form of authoritative verification. As these tools become more accessible, the tendency to treat AI generated text as an objective truth continues to create friction between technical capability and public accountability standards.