Jenia Iontcheva Turner (Southern Methodist University – Dedman School of Law) has posted AI Discovery in Criminal Cases (American Criminal Law Review, forthcoming) on SSRN. Here is the abstract:
The growing complexity and volume of digital data threaten to overwhelm criminal law practice. Prosecutors and defense attorneys are struggling to process and review gigabytes of video, audio, and text files that now accompany even run-of-the-mill criminal cases. Policymakers and courts have failed to offer adequate remedies for this growing problem, so many practitioners have turned their attention to AI as a potential solution. AI tools can offer convenience and time savings for prosecutors and defense attorneys managing massive digital evidence volumes. Yet the benefits delivered by the technology sometimes come at the cost of less accurate analysis, loss of client confidentiality, and unfairness in the criminal process.
This Article presents the results of an original survey analyzing how prosecutors and defense attorneys approach the risks and benefits of AI tools in legal practice. The survey responses suggest that criminal law practitioners use AI cautiously and selectively, for tasks where the perceived risk of problems is low or can be minimized through human review. This careful approach reflects attorneys’ appreciation of the constitutional and professional consequences of AI errors. If a prosecutor uncritically relies on AI to review voluminous materials for exculpatory evidence, and AI fails to find it, this can violate a defendant’s right to due process and result in a wrongful conviction. Likewise, a defense attorney who fails to uncover evidence favorable to the client because of undue deference to AI tools may be providing constitutionally ineffective assistance of counsel. In addition, misusing AI tools can subject both prosecutors and defense attorneys to professional discipline.
Recognizing these risks and the constitutional rights at stake, the Article suggests measures that lawyers can implement to guide the use of AI in criminal case discovery. The Article also argues for stricter regulation and public investment to ensure that AI discovery tools are reliable, secure, and broadly accessible to criminal law practitioners.
Recommended!
To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.
Lawrence Solum
