Skip to main content

Blog · Legal Ethics

The Duty of Technology Competence: Comment 8, the States, and What It Means for AI

Since 2012, ABA Model Rule 1.1 has told lawyers to keep up with the benefits and risks of relevant technology. Most states have adopted some version. Here is the exact text, a sourced count of the adopting jurisdictions and how their versions differ, and what the duty requires once generative AI is in the office.

The duty of technology competence comes from Comment 8 to ABA Model Rule 1.1, added in 2012: a lawyer should keep abreast of "the benefits and risks associated with relevant technology." According to the LawSites tracker, 40 states plus D.C. and Puerto Rico have adopted some version. For AI, it means understanding a tool's limits before relying on it and checking its output.

The duty sat in the background for a decade, applied mostly to e-discovery, metadata, and cloud storage. Generative AI moved it to the front, because a lawyer who does not understand how a chatbot produces citations is exactly the risk the comment describes. This post gives the text, a sourced count of where it applies and how the versions differ, and what it requires in an office that uses AI.

The text of Comment 8

Rule 1.1 itself is one sentence: "A lawyer shall provide competent representation to a client," requiring "the legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation." The technology duty lives in the comment, under the heading "Maintaining Competence." The current ABA text reads:

[8] To maintain the requisite knowledge and skill, a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology, engage in continuing study and education and comply with all continuing legal education requirements to which the lawyer is subject.

The ABA House of Delegates added the words "including the benefits and risks associated with relevant technology" in August 2012, on the recommendation of the ABA Commission on Ethics 20/20. The same package amended Rule 5.3, retitling it "Responsibilities Regarding Nonlawyer Assistance," which matters for AI too.

Three things about the wording are worth noticing:

  • "Benefits and risks." The duty runs both ways. It is not only about avoiding dangerous tools; it is also about knowing what useful tools exist.
  • "Relevant." It covers the technology relevant to your practice, not all technology. A probate lawyer and an e-discovery specialist have different obligations.
  • "Should." Comments explain rules rather than impose separate duties. In practice the comment shapes what "competent representation" in Rule 1.1 means, and that rule is mandatory.

Which states have adopted it

A Model Rule comment binds no one until a state's high court or rule-making body adopts it. Bob Ambrogi's LawSites has tracked adoptions since 2012, and as of November 2026 its tech competence page lists 40 states, the District of Columbia, and Puerto Rico. The last states added were Hawaii, effective January 1, 2022, which LawSites reported brought the total to 40, followed by D.C. and Puerto Rico in 2025.

Counts you see elsewhere vary because the versions vary. The notable differences:

JurisdictionWhat it adoptedSource
Most adopting statesThe ABA's Comment 8 language, verbatim or close to itLawSites tracker
North CarolinaSlightly different wording: "the technology relevant to the lawyer's practice"NC 2024 FEO 1 quoting its Comment 8
HawaiiAmended Comment 6 instead of Comment 8 and added technology language to a Rule 5.3 commentLawSites, 2022
District of ColumbiaApproved April 7, 2025; added "technology" to Comment 5 (thoroughness on a particular matter), not Comment 8LawSites, 2025
Puerto RicoA standalone Rule 1.19, "Technological Competence and Diligence," effective January 1, 2026, with comments that address AILawSites, 2025
New JerseyDeclined to adopt the duty in 2025 but adopted a technology CLE requirementLawSites, 2025

Utah adopted the duty in March 2015, per the tracker. Our guide to Utah AI ethics for lawyers covers how it interacts with Utah's other AI authorities.

Two caveats. First, a state that has not adopted the comment still requires competence; the D.C. Bar, before its 2025 change, said competence "always requires the legal knowledge, skill, thoroughness, and preparation reasonably necessary" and applied that to technology anyway. Second, the tracker is maintained by a journalist, not an official body. For a specific state, read that state's current Rule 1.1 and comments.

CLE: where the duty has hours attached

A few jurisdictions turned the comment into a training requirement. The Florida Bar was the first mandatory bar to require technology CLE, with three hours of technology programs per reporting cycle, as the Florida Bar News reported. North Carolina became the second, requiring one hour of technology training starting in 2019, per LawSites. New Jersey, as noted, chose the CLE route instead of the comment. If your state has such a requirement, an AI course is the obvious way to meet it; check your bar's current rules for the hours.

What the duty means once AI is in the office

The duty predates generative AI, but the ethics opinions on AI now define what it requires. ABA Formal Opinion 512 (July 29, 2024) is the clearest statement, and its competence section reads like a gloss on Comment 8.

A reasonable understanding, not expertise. Lawyers "need not become GAI experts" but must have "a reasonable understanding of the capabilities and limitations of the specific GAI technology that the lawyer might use." Note "specific": the duty attaches to the tool you use, not to AI in general.

Continuing, not one-time. "This is not a static undertaking." The opinion suggests reading about tools aimed at the profession, attending CLE, and consulting people who know the technology. A tool that worked one way when you adopted it may work differently after an update.

Verification proportional to the task. The opinion says the required review "will necessarily depend on the GAI tool and the specific task." Its example: a lawyer who tested a tool's contract summaries against a hand-reviewed subset and found them accurate need not manually review every contract. It also says a tool "designed specifically for the practice of law" may require less independent review, particularly where the lawyer's experience gives a reasonable basis for relying on it.

No abdication. Lawyers "may not abdicate their responsibilities by relying solely on a GAI tool to perform tasks that call for the exercise of professional judgment." Whatever level of review you choose, "the lawyer is fully responsible for the work on behalf of the client."

Maybe, someday, a duty to use it. The opinion says "it is conceivable that lawyers will eventually have to use them to competently complete certain tasks for clients," while stopping short of saying that time has come. For now, lawyers "should become aware of the GAI tools relevant to their work so that they can make an informed decision" whether to use them. D.C.'s Opinion 388 makes the same point: "there may come a time when lawyers' use of GAI is standard practice."

That last point is the "benefits" half of the comment, and it is easy to forget. Technology competence is not satisfied by refusing to touch AI. It is satisfied by knowing enough to make an informed choice. Our post on whether lawyers can use AI for legal research covers that choice in more depth.

A working standard for a small firm

Put together, the comment and the AI opinions suggest a routine that is achievable without an IT department:

  1. Know how each tool gets its answers. The single most useful question about an AI research tool is whether it retrieves real sources and cites what it retrieved, or generates text from a model's memory. The D.C. Bar drew the same line, saying less diligence may be warranted for tools that "rely on proven, current databases." Our guide to AI legal research accuracy explains how to tell.
  2. Test on known answers. Before relying on a tool, give it a question from a matter you have already researched and compare.
  3. Review in proportion to stakes. A brainstorming prompt needs less review than a brief. Every citation in a filing needs checking; our citation verification guide covers the method.
  4. Recheck periodically. Put a date on the calendar to reread each vendor's terms and release notes.
  5. Train the people who use it. Supervising lawyers answer for staff use under Rules 5.1 and 5.3. A written AI policy and one training hour cover most small firms.
  6. Document the choice. If you decide a tool is not reliable enough for a task, write down why. That record is competence too.

Item 1 is the design choice behind CaseRead: research that searches your matter files and public law together and cites only sources it actually returned. The features page describes how that works. For the jurisdiction-specific AI opinions that sit on top of the competence duty, see our map of state bar AI guidance.

Frequently asked questions

What is the duty of technology competence?

It is the part of a lawyer's duty of competence under Rule 1.1 that covers technology. Since 2012, Comment 8 to ABA Model Rule 1.1 has said a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology. It applies in a state once that state's high court adopts it.

How many states have adopted the duty of technology competence?

LawSites, which has tracked adoptions since 2012, counts 40 states plus the District of Columbia and Puerto Rico as of November 2026. Versions differ: D.C. put technology in a different comment, Hawaii amended a different comment and added language to Rule 5.3, and Puerto Rico adopted a standalone rule effective January 1, 2026. New Jersey declined in 2025.

Does the duty of technology competence require lawyers to use AI?

Not today. ABA Formal Opinion 512 says it is conceivable lawyers will eventually have to use generative AI for certain tasks, but for now they should know the tools relevant to their work well enough to decide whether to use them. If they do use one, they must understand its capabilities and limits and review its output.

Is there a CLE requirement for technology?

In some states. Florida was the first mandatory bar to require technology CLE, with three hours per reporting cycle, and North Carolina began requiring one hour of technology training in 2019. New Jersey declined to adopt the competence comment in 2025 but adopted a technology CLE requirement. Check your own state's CLE rules for current hours.

What does technology competence mean for AI legal research?

It means knowing how the specific tool produces its answers, testing it on work whose answer you know, and reviewing its output in proportion to the risk. ABA Formal Opinion 512 calls this a reasonable understanding of the tool's capabilities and limitations and warns it is not a static undertaking, because the tools keep changing.

A competence check you can run today

The fastest practical test of technology competence is whether the citations in your last AI-assisted draft are real and say what the draft claims. Paste it or upload the file into CaseRead's free Hallucination Shield: it checks up to 50 citations per run against official statute and case-law sources, with no signup.