The DOJ has put companies on fair notice of its expectations on how companies can responsibly adopt AI tools. Companies that take the necessary steps may benefit from a de facto “compliance defense, but those companies that fail to heed this guidance expose themselves to the risk of prosecution.
Artificial intelligence (AI) continues to dominate the headlines as a transformational tool that promises to change our lives in fundamental ways. Companies in particular are increasingly including AI tools as part of their strategic plans to make their business operations more productive, efficient, and ultimately, profitable. A subset of those companies also are incorporating AI into customer-facing solutions, and as investments continue to flow into AI, the pace of adoption will only increase.
At this relatively early stage, however, it is not readily clear what threats AI may pose to society, including potential violations of criminal laws. As we explained in an earlier article, because an AI system has no mind with which to form a criminal intent, where decisions made by AI systems lead companies to violate the law, prosecutors will likely face substantial hurdles under existing laws to prosecute corporations for crimes caused by faulty AI decision-making. But that does not mean that prosecutors are powerless to charge corporations with AI-based crimes. A corporation may still be charged if it fails to take adequate steps to test and control the AI system to prevent the violations from happening in the first place. Companies that fail to do so could face charges for compliance-based crimes. Notably, these compliance-based crimes should not be seen as second-class citizens, as evidenced by recent billion-dollar corporate resolutions brought by the Department of Justice (DOJ) charging what are essentially compliance-based failures, such as failing to implement an effective system of internal accounting controls or failing to adopt an adequate anti-money laundering program. As a general operating principle, companies should assume that their compliance programs will, at some point, become the focus of a criminal investigation. For companies adopting AI solutions for their business operations, that principle takes on heightened importance.
