ABA Opinion 512 Sets Clear AI Ethics Rules for Legal Professionals

The ABA issued Formal Opinion 512 on July 29, 2024, establishing ethics guidance for AI use in law.

The American Bar Association has issued its Formal Opinion 512 on July 29, 2024, clarifying that all core ethical rules—confidentiality, competence, and supervision—squarely apply when lawyers deploy artificial intelligence in client work.

• AI-powered document review, research assistants, and basic brief drafting are now in use at 42% of legal service providers, a jump from 26% last year.
• Yet, Opinion 512 stresses legal professionals cannot delegate ethical duties to technology vendors or ignore the workings of "generative AI" (systems that autonomously create text or data). Ongoing oversight is mandated.
• Lawyers are urged to evaluate each AI tool’s risk of client information disclosure, unauthorized access, and the strength of security. The ABA advises ensuring contractual data protections and in-firm control over sensitive data.

Visibility gap: One persistent risk, highlighted by a recent survey, is in-house teams’ low awareness of vendors’ or outside counsels’ AI adoption—60% say they do not know whether their counsel uses generative AI on their matters.

For risk and compliance teams, the Opinion signals that AI-enabled efficiency cannot come at the expense of diligence: rigorous vetting, training, and clear client communications are all required. As Nicholas Daniel Seger notes in the ABA’s practical implementation guide, the best practice is a cautious rollout with direct attention to who manages the data and how confidentiality is preserved.