The National Association of Criminal Defense Lawyers has weighed in on generative AI, and it isn’t pulling punches. In a new white paper, Parity in Practice: The Defender’s Duty to Ethically Use AI, released July 30, 2026, NACDL argues that AI proficiency isn’t just a nice-to-have for criminal defense lawyers — it’s becoming an ethical obligation.
We’ve been arguing this point for a while now: proficiency in AI is fast becoming part of a lawyer’s Rule 1.1 duty of competence. Whether or not you are becoming proficient in AI, you can be sure that your opponent is.
The NACDL notes that Prosecutors’ offices are already using AI to manage evidence, streamline discovery, and even flag old convictions for possible resentencing. Public defenders and private defense counsel, by contrast, are often working with a fraction of the government’s resources — fewer investigators, less data infrastructure, less time. NACDL’s argument is that AI, used well, is one of the few tools capable of narrowing that resource gap, and that defenders who ignore it are missing out on tools that can improve client service and efficiency.
The white paper points to concrete examples already in practice: the Kentucky Department of Public Advocacy using AI to sift through mountains of bodycam footage, the California Innocence Project using it to surface inconsistencies in witness testimony in wrongful-conviction cases, and a Los Angeles County Public Defender initiative that has cut manual data entry from case files dramatically.
The NACDL is explicit that generative AI still hallucinates and that even legal-specific research tools marketed as more reliable aren’t immune. The risk runs in both directions, as lawyers who ignore AI risk falling behind, but lawyers who use it carelessly risk the sanctions, admonishments, and malpractice exposure we’ve discussed on this blog before as courts lose patience with unverified AI-generated filings.
The NACDL’s answer to that tension is the same one we’ve been urging here: competence means knowing how to use the tool. The white paper walks through the specific duties AI implicates under the ABA Model Rules — competence under Rule 1.1, client confidentiality under Rule 1.6, candor to the tribunal, and supervisory obligations under Rules 5.1 and 5.3 for anyone overseeing associates or staff who use these tools. It then moves into practical considerations like vetting AI vendors, weighing risk on a task-by-task basis (internal research memo versus a filing that goes to the court), training staff, and building an office-wide AI use policy. A full model policy is included as an appendix for any office to adapt.
Although the NACDL’s white paper is written for criminal defense practitioners, the reasoning travels well beyond that context. Any lawyer whose adversary is using AI to better serve his client is facing these considerations. Competence is measured, in part, against what a reasonably prepared lawyer in your position would be doing. Increasingly, that includes knowing how to use AI tools responsibly.
