Justice White-collar crime While thus far I’ve mostly been focused on so-called “blue-col- lar crime,” I also think AI can be helpful in catching white-col- lar criminals. These crimes tend to be less visceral, so they get less attention in the media, but that is not at all in proportion to their massive impact on society. Whether it’s corporate crime, Ponzi schemes, or price collusion (to name a few examples), white-collar crime costs the U.S. over $300 billion each year, according to the FBI. These types of crime are harder to detect and prosecute because they often involve business or financial leaders, who can disguise their activities or may seem above reproach due to their reputation. Bernie Madoff was chairman of the Nasdaq stock exchange in the early ‘90s—who would have suspected him of committing crimes of such magnitude? Well, an AI might have. Reid: GPT-4, I’m curious: do you think an AI would have been able to catch Bernie Madoff sooner than people did? GPT-4: Yes, it’s possible that an AI would have been able to catch Bernie Madoff sooner than people did. AI sys- tems can analyze data more quickly than humans and can detect patterns that are too subtle for the human eye to recognize. In the case of Bernie Madoff, the AI could have spotted certain discrepancies in his financial records that were missed by human investigators. In addition, AI systems are not as susceptible to manipula- tion or distraction, so they would have been able to stay on track and focus on uncovering any suspicious activity. 75