Ellen S. Podgor (Stetson University – College of Law) has posted Smart Crime on SSRN. Here is the abstract:
As Artificial Intelligence (AI) grows increasingly sophisticated and autonomous, the lack of a criminal law framework to deter misconduct poses unprecedented challenges. Not surprisingly, the historical development of the computer crimes statute faced similar roadblocks when arguments were made that a specific computer crime statute was unnecessary and existing legislation would suffice. Shoehorning computer-related misconduct into existing statutes—espionage, stalking, wire fraud—failed as an approach, prompting Congress to eventually pass the Computer Fraud and Abuse Act.
Equally daunting is the effort to criminalize AI-related misconduct through a traditional Model Penal Code framework—it just doesn’t work. Simple questions raise unprecedented issues in the AI context. For example, who committed the voluntary act? Is it the developer, the coder, the user, or the algorithm itself?
This Article considers the current tensions between state and federal approaches to AI regulation, surveys the landscape of international criminal law, and considers comparative AI criminal law. It calls for a recalibration of classic criminal law principles to honor normative values in a new and fast developing AI world. Not only does this Article demonstrate how existing laws do not sufficiently address AI crimes, it also offers a blueprint for how lawmakers should craft AI criminal legislation that protects society without stifling creative innovation, particularly innovation that may drive progress toward Artificial General Intelligence (AGI). Most importantly, this needs to happen sooner rather than later. History has taught us the cost of legislative delay in the belated arrival of the computer crime statute. AI criminal law must be proactive, comprehensive, and a full panoply of stakeholders needs to be at the drafting table.
Highly recommended.