Hawaii AI Ethics Rules for Lawyers

The official instruments

A committee the chief justice built

Hawaii’s answer to generative AI arrived as a judiciary project with a docket number. Chief Justice Mark E. Recktenwald established the Committee on Artificial Intelligence and the Courts on April 16, 2024, placed Associate Justice Vladimir P. Devens and Judge John M. Tonaki at its head as co-chairs, and the committee’s Final Report to the Hawaii Supreme Court, dated December 15, 2025 and e-filed December 16, 2025, now sits in the court’s filing system under SCMF-24-0000305. The bar side of the state stayed quiet in the meantime: we inspected the Disciplinary Board’s formal-opinions index on July 1, 2026 and found no AI opinion there, so the report is Hawaii’s operative instrument. The tracker types it as what it is, a report, with day precision taken from the document’s own face, per the labeling rules on the methodology page.

The ethics conclusion: the rules already reach it

The report’s central professional-responsibility holding comes from its Ethics Subcommittee: existing rules, including the Hawaii Rules of Professional Conduct, are broad enough to cover generative AI. On that conclusion the committee proposed no rule changes at all, describing its stance as wait and see. Hawaii thereby joins the restraint camp, the jurisdictions that read their current rulebooks as already adequate, and the closest structural parallel in the dataset is Vermont, whose judiciary committee declined to draft amendments by a similar route: supreme court creates a committee, the committee studies, the committee chooses guidance over text. The consequence for Hawaii lawyers is that nothing AI-specific will appear in the HRPC for now; competence, confidentiality, candor, and supervision do the work in their existing form, with ABA Formal Opinion 512 as the fullest articulation of how those duties run through generative AI practice.

What the committee asked for instead

Three recommendations carry the report’s weight. Education comes first, at every level of the profession: continuing legal education for practicing lawyers, AI ethics teaching in the law schools, and judicial education for the bench. Second, a standing judiciary committee to vet AI tools, which would give Hawaii a permanent institutional checkpoint between new technology and courtroom use rather than ad hoc adoption decisions. Third, continuation of the committee itself, keeping a body in place to watch a technology that moves faster than reporting cycles. The report also turns the lens inward and surveys the judiciary’s own AI use, including the KolokoloChat chatbot and the internal guardrails set out in Appendix 3. A court system that documents its own tooling while declining to regulate the profession’s is making a considered bet: build capability now, write rules only if the watching reveals a need.

Practicing in a wait-and-see jurisdiction

For a Hawaii firm the compliance work is the national baseline, executed well, because there is no local overlay to memorize. Tool vetting comes before any client information is exposed, and the risk-tier framework gives that decision a structure. Verification of outputs comes before use, a discipline the hallucination sanctions guide exists to motivate. Billing tracks actual time, as the billing analysis explains, and a written office policy makes supervision demonstrable; the policy template assembles those parts into one document. Filing obligations remain a judge-by-judge question covered in the court orders guide. What makes Hawaii distinctive is that its instrument names its own successors, an education program, a vetting committee, a continuing body, and each is a concrete future event; the changelog records each one the week it verifies.