Technology-Competence Duty Expanded: Rhode Island Adds “Benefits and Risks” of Developing Technology to Rule 1.1 and Issues Interim Generative AI Ethics Guidance

1. Introduction

In In re Amendments to Article V, Rule 1.1 of the Supreme Court Rules (Professional Conduct) and Adoption of Interim Generative AI Guidelines (R.I. June 17, 2026), the Supreme Court of Rhode Island issued an Order that (1) amends Comment [6] to Rule 1.1 (Competence) of the Rhode Island Rules of Professional Conduct, and (2) adopts two sets of “Interim Guidelines” addressing the ethical use of generative artificial intelligence (“GAI”) by Rhode Island lawyers and by Rhode Island judicial officers.

The key issues are forward-looking and regulatory rather than adjudicative: how professional “competence” should be defined in an era where generative systems can draft text and produce outputs that appear authoritative; and how lawyers and judges can use GAI consistently with long-standing duties such as confidentiality, candor, impartiality, and independent judgment.

The parties are not litigants in a conventional sense. This is a court-administered change to governing professional standards (Article V and related guidance) and an advisory ethical framework for the bench (Article VI guidance).

2. Summary of the Opinion (Order)

2.1 Amendment to Rule 1.1 Comment [6]

The Court amends Comment [6] to Rule 1.1 to clarify that maintaining competence includes keeping abreast of changes in law and practice “including the benefits and risks associated with existing and developing technology.” This language places technology awareness squarely within the competency framework, without mandating adoption of any particular tool.

2.2 Adoption of Interim GAI Guidelines (Advisory, Not Binding)

The Court adopts attached “Interim Guidelines for the Ethical Use of Generative AI by Rhode Island Lawyers” and “Interim Guidelines for the Ethical Use of Generative AI by Rhode Island Judicial Officers,” while expressly stating:

  • They “neither amend” Article V (Rules of Professional Conduct) nor Article VI (Code of Judicial Conduct).
  • They “nor carry the force of law.”
  • They are “advisory in nature” and subject to revision as GAI evolves.

The practical consequence is that the amendment changes the competency baseline, while the Guidelines operate as an authoritative ethical roadmap for compliance with existing rules in common GAI scenarios.

3. Analysis

3.1 Precedents Cited

The Order and attached Guidelines do not cite adjudicated judicial precedents (no prior case titles are referenced). Instead, they rely on and interpret the text and Comments of:

  • Article V: Rhode Island Rules of Professional Conduct (e.g., Rules 1.1, 1.3, 1.4, 1.5, 1.6, 1.7, 1.10, 1.18, 2.1, 3.1, 3.3, 5.1, 5.3, 7.1–7.5, 8.3, 8.4; and Preamble [10]).
  • Article VI: Rhode Island Code of Judicial Conduct (e.g., Rules 1.2, 2.2, 2.3, 2.4, 2.5, 2.9, 2.12, 2.15; and referenced Comments).

In this sense, the “precedential” force of the Order comes not from common-law reasoning across prior cases, but from the Court’s constitutional and supervisory authority over the practice of law and judicial ethics—implemented through rule amendment (binding) and interpretive guidance (advisory) that anticipates disciplinary and professional-responsibility questions.

3.2 Legal Reasoning

A. Competence as Technology-Risk Management (Lawyers)

The amendment to Comment [6] and the Guidelines’ discussion of Rule 1.1 treat technology literacy as part of “the legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation.” The Court’s conceptual move is to treat GAI not as a novelty outside the Rules, but as another practice-changing tool whose “benefits and risks” must be understood for a lawyer to remain competent.

Importantly, the Guidelines draw a boundary: the duty is awareness and informed use, not compulsory adoption. A lawyer need not use GAI, but must understand it sufficiently to avoid incompetent representation where GAI is used by the lawyer, the firm, opposing counsel, or becomes a feature of modern legal practice (the Guidelines analogize to email’s rapid normalization).

B. Diligence and the Non-Delegation of Professional Responsibility

Under Rule 1.3 (and, for judges, Rule 2.5), the Guidelines emphasize that efficiency gains do not justify “outsourcing” core professional tasks. The lawyer remains accountable for the work product. This is operationalized most sharply in litigation duties: because GAI can “hallucinate,” the lawyer must verify citations, facts, and legal statements before filing (Rules 3.1 and 3.3), and must correct errors to avoid misrepresentation (including unintentional misstatements).

C. Communication: When GAI Use Becomes Material to the Client

The Guidelines interpret Rule 1.4 as a materiality-driven disclosure regime. Lawyers are not required to disclose every use of GAI, but should communicate about GAI use when doing so advances the purposes of Rule 1.4—i.e., when it affects the means of representation in a way that informs client participation and decision-making. The Guidelines identify factors such as significance to the task, importance to the representation, and how disclosure affects the client’s understanding and confidence.

Two disclosure triggers are treated as especially important:

  • When a client asks about GAI use.
  • When client information is entered into a GAI system (linked to Rule 1.6 confidentiality concerns).

D. Fees: GAI Efficiency and the Meaning of “Time Spent”

The Guidelines treat Rule 1.5 as prohibiting “phantom billing” based on what a task used to take without GAI. If GAI reduces time, the time billed must reflect actual time spent. In addition, routine GAI subscriptions used for everyday work are framed as office overhead, not reimbursable client expenses—analogized to rent, utilities, productivity software, and insurance.

The document preserves limited room for client-charged costs: specialty tools not regularly used may be charged as reasonable out-of-pocket costs when requested by the client or necessitated by the representation, consistent with the general expense rubric in Rule 1.5’s Comments.

It also draws a training distinction:

  • General GAI learning is part of competence and not client-billable.
  • Learning a specific tool requested by the client may be chargeable as a reasonable cost in furtherance of the representation.

E. Confidentiality and Tool Governance

The Guidelines operationalize Rule 1.6 through a “know your vendor/tool” approach: lawyers should review Terms of Use and privacy/security policies to determine whether data is stored, reused, disclosed, or retained after service termination; whether breach notice is provided; and whether confidentiality obligations are enforceable. The open/closed model distinction matters because some tools may retain prompts and outputs or use them for training, creating a confidentiality and security risk.

The Guidelines implicitly position vendor diligence and security assessment (sometimes requiring consultation with technology professionals) as a component of ethical compliance—not merely IT best practice—because careless tool selection can constitute an ethical breach if client information is exposed.

F. Conflicts, Screening, and Firm-Wide Systems

Under Rules 1.7 and 1.10, the Guidelines highlight a modern screening problem: even if a firm screens a conflicted lawyer, GAI-enabled internal tools must also be configured so the screened lawyer cannot access restricted client information “inadvertently or otherwise.” This extends the screening concept from file cabinets and document management systems to AI-assisted knowledge bases, chat interfaces, and shared prompt/output repositories.

G. Prospective Clients and GAI Chatbots

The Guidelines treat website chatbots and “agents” as a risk point under Rule 1.18, emphasizing that the impersonal medium does not reduce duties. Potential pitfalls include unintended receipt of confidential information, misstatements about services, and inadvertent formation of an attorney-client relationship. The message is that intake automation is ethically consequential and must be designed and supervised accordingly.

H. Independent Professional Judgment

Under Rule 2.1, the Guidelines allow GAI as a consultative aid—analogous to treatises, research databases, or experts—but reject substitution of GAI output for the lawyer’s own candid advice. This frames GAI as a tool for generating possibilities, not the authoritative source of judgment.

I. Candor, Meritorious Claims, and “Hallucinations”

The Guidelines repeatedly return to the litigation risk: GAI can produce plausible but false citations and analysis. The lawyer’s ethical obligations under Rules 3.1 and 3.3 (and related Rules 4.1 and 8.4(c)) require verification and correction. Notably, the Guidelines state these prohibitions apply to intentional and unintentional statements and actions, underscoring that “I relied on the tool” is not a defense to filing falsehoods.

J. Supervision: Institutionalizing Compliance

Rules 5.1 and 5.3 are used to place responsibility on partners/managers for firm-wide GAI governance: policies on permissible use, training on confidentiality and security, and accountability mechanisms for employees and contractors (including contractual safeguards and forum/recourse).

K. Advertising and Unauthorized Autopublishing

Rules 7.1 through 7.5 are applied to AI-generated marketing: lawyers remain responsible for accuracy and non-misleading communications even where content is produced or placed with minimal human involvement.

L. Self-Regulation and Reporting Misconduct

The Guidelines integrate GAI into the profession’s self-policing norms (Rules 8.3 and 8.4; Preamble [10]), warning that sanctions have been imposed elsewhere for GAI misuse and that lawyers must exercise diligence regarding opposing counsel’s outputs. The key interpretive point is the definition of “substantial” in Rule 8.3’s commentary: seriousness, not quantity of evidence; “even a single incorrect citation” may warrant action in appropriate circumstances.

M. Judicial Officers: Independence, Bias, Ex Parte Risks, and Drafting Limits

The judicial Guidelines map GAI use onto the Code’s core themes:

  • Rule 1.2 (Confidence in the Judiciary): avoid impropriety and appearance of impropriety in GAI use.
  • Rule 2.2 (Impartiality and Fairness): hallucinations and inadequate training may produce errors that may not be “good-faith errors” protected by Comment [3].
  • Rule 2.3 (Bias, Prejudice, Harassment): GAI may encode stereotypes; judges must filter outputs through their own training and ensure staff/litigants do not deploy biased content.
  • Rule 2.4 (External Influences): GAI-generated authoritative-seeming text is framed as a potential “undue influence.” The Guidelines state judges “may consult or review but should never rely on GAI-produced content” when drafting orders and decisions.
  • Rule 2.9 (Ex Parte Communications / Independent Fact Investigation): GAI outputs can import facts beyond the record; judges must ensure fidelity to the record and take remedial steps when outside information appears.
  • Rules 2.12 and 2.15 (Supervision and Responding to Misconduct): judges should enact policies for staff GAI use and take appropriate action regarding lawyer/judicial misconduct related to GAI.

The most consequential normative line is the drafting caution under Rule 2.4: the Guidelines treat reliance on GAI-drafted decision text as a risk to perceived independence and to actual decision integrity, even if the judge believes the tool is merely “helping.” The emphasis is on the non-delegable nature of judicial reasoning and on maintaining public confidence.

3.3 Impact

A. Immediate Doctrinal/Regulatory Effect

  • Binding change: Comment [6] to Rule 1.1 now expressly includes “benefits and risks” of technology within the competence framework.
  • Persuasive guidance: the Interim Guidelines, while not law, are likely to be treated as influential benchmarks for what constitutes “reasonable” conduct under existing Rules and Code provisions.

B. Practice Management Consequences for Lawyers

  • Verification becomes standard-of-care adjacent: especially for citations, quotations, and fact assertions in filings.
  • Vendor/tool due diligence: Terms of Use, privacy, retention, breach notice, and training-use policies become ethics-relevant.
  • Billing reforms: pressure against billing based on pre-GAI time expectations; clearer overhead vs reimbursable expense distinctions.
  • Documented client communication: more engagement letters may include GAI disclosures and data-handling choices.
  • Firm governance: written policies, training, access controls, and audit trails for AI tools may become routine to satisfy supervisory obligations.

C. Consequences for Courts and Judicial Administration

  • Internal policies: chief/admin judges may issue global orders on staff use, consistent with Rule 2.12 guidance.
  • Litigant AI use management: judges may need to balance access and fairness with courtroom integrity, resource constraints, and rights of opposing parties.
  • Record integrity: heightened vigilance against extra-record information entering judicial reasoning via AI summaries or searches.

D. Disciplinary and Sanctions Trajectory

By connecting hallucination risks to Rules 3.1, 3.3, 8.3, and 8.4, the Guidelines foreshadow increased scrutiny of inaccurate filings, including the possibility that failures to verify may be framed as competence/diligence violations even absent intent to deceive.

4. Complex Concepts Simplified

  • “Hallucination”: when a GAI tool generates content that appears credible but is false (e.g., fictitious case citations or invented quotations).
  • “Informed consent” (confidentiality context): client agreement after being given enough information to understand what will be disclosed and the risks of disclosure (including, here, risks tied to a tool’s data retention and reuse).
  • “Impliedly authorized” disclosure: limited sharing that is reasonably necessary to carry out representation; the Guidelines caution that many GAI tools’ data practices may exceed what is “reasonably necessary.”
  • “Screening”: firm procedures that isolate a conflicted lawyer from access to protected information; the Guidelines extend this to AI systems so screened lawyers cannot retrieve restricted data through shared tools.
  • “Office overhead” vs “client expense”: routine operational costs (including everyday GAI subscriptions) should not be billed as client expenses; specialty, matter-specific tools may sometimes be charged if reasonable and properly agreed.
  • “Ex parte” and “independent fact investigation” (judges): judges generally may not gather facts outside the record; GAI outputs can inadvertently introduce outside information, so judges must ensure outputs stay within the record and permissible judicial notice.
  • “Appearance of impropriety”: even if conduct is not actually biased or improper, it can violate judicial ethics if it reasonably undermines public confidence in independence and impartiality.

5. Conclusion

This Order does two important things for Rhode Island’s legal system. First, it formally embeds technology risk/benefit awareness into the concept of professional competence by amending Comment [6] to Rule 1.1. Second, it supplies an extensive, rule-anchored set of Interim Guidelines that translate traditional duties—competence, diligence, confidentiality, candor, supervision, impartiality, and independence—into concrete expectations for GAI use.

The overarching takeaway is continuity with sharper edges: GAI does not change ethical duties, but it increases the frequency and severity of ways those duties can be breached (fabricated authority, leaked data, hidden extra-record facts, biased outputs, and misleading communications). Lawyers and judges may use GAI, but they must do so in a way that preserves human responsibility for judgment, accuracy, confidentiality, and fairness.