Defense attorneys have an ethical obligation to responsibly adopt certain artificial intelligence tools, similar to prosecutors who increasingly use this technology. A report by the National Association of Criminal Defense Lawyers outlines the advantages and disadvantages of using current AI tools. The recommendation is for defenders to explore "enterprise" AI tools, such as Microsoft Copilot via a Microsoft license, or legal-specific tools like Harvey and Thomson Reuters’ CoCounsel, instead of commercial products like ChatGPT, Gemini, or Claude due to confidentiality concerns.

The focus on restricting the use of general commercial AI in favor of enterprise or legally vetted tools reveals a tension between technological capability and professional responsibility. The necessity for defense counsel to handle sensitive client information elevates the risk profile associated with off-the-shelf models, suggesting that confidentiality concerns are not merely logistical issues but ethical determinants. This signals a pattern where emerging technology introduces complexity regarding fiduciary duty, forcing practitioners to navigate the divide between efficiency gains and legal safeguards. The implication is that the adoption of AI in the legal sphere must be framed not just by capability but by an assessment of risk management specific to adversarial contexts. What are the systemic implications for defining professional competence when the tools used—and the data they process—carry differential privacy risks? How does the established framework of "ethical obligation" translate into actionable, scalable policies for technological adoption across the legal profession?