Oregon's two AI ethics opinions, explained

Informational only, not legal advice. TacticalMS is not a law firm. Read more.

Oregon lawyers now have two formal ethics opinions from the Oregon State Bar that deal directly with artificial intelligence. The first, Formal Opinion 2025-205, was approved by the Board of Governors in February 2025 and looks at AI tools in law practice generally ¹. The second, Formal Opinion 2026-208, was approved in February 2026 and focuses on chatbots and AI agents that communicate with clients and prospective clients ².

Formal opinions are the bar's reading of how the Oregon Rules of Professional Conduct apply to a situation. They are guidance, not new rules, but they are the clearest statement available of how the bar views these tools. This post summarizes what each opinion says. It is not a substitute for reading them, and both are linked below.

Opinion 2025-205: AI tools in practice

The 2025 opinion is broad. A Northwest ethics commentator described it as similar in scope to ABA Formal Opinion 512, the American Bar Association's 2024 opinion on generative AI ³. Its main points, as summarized from the opinion itself ¹:

• Competence (RPC 1.1). A lawyer may use AI only after taking reasonable steps to understand its specific benefits and risks, including how the tool stores information and whether it uses prompts for training. The opinion treats this as an ongoing duty ¹.
• Confidentiality (RPC 1.6). The opinion distinguishes "open" models that keep learning from what users enter from "closed" models that keep data in a secure environment. Putting client information into an open model calls for informed consent or complete anonymization. Closed models still call for a review of the vendor's terms on storage, encryption and deletion ¹.
• Fees (RPC 1.5). Lawyers may bill reasonable time for case-specific work done with AI, but not in a way that inflates hours. Charging extra for work already covered by a flat fee raises excessive-fee concerns, and passing actual AI costs to a client should be disclosed in advance, preferably in writing. Time spent learning a tool is generally not billable ¹.
• Supervision (RPC 5.1 to 5.3). Firm managers are expected to set AI policies and train lawyers and staff on the specific tools in use ¹.
• Candor (RPC 3.3 and 4.1). Lawyers must verify AI output, especially citations, quotations and case-specific facts, and must check court rules on AI use ¹.

Opinion 2026-208: chatbots and AI agents

The 2026 opinion asks a narrower question: may Oregon lawyers use chatbots to communicate with clients or prospective clients on their websites or elsewhere? Its answer is "Yes, qualified" ². Our plain-English guide to Opinion 2026-208 goes through it rule by rule. The qualifications track the same rules, applied to a tool that talks to the public on the firm's behalf:

• Prospective clients (RPC 1.18). Information a prospective client shares with a chatbot is protected the same way as information shared with a lawyer, and it can create conflicts. The opinion suggests entering chatbot information promptly into the firm's conflict system ².
• Accidental attorney-client relationships. Because Oregon looks at the client's reasonable belief, a chatbot that sounds like a lawyer could create a relationship the firm didn't intend. The opinion points to clear disclaimers and to having a responsible lawyer notified of chatbot conversations ².
• Saying it's an AI (RPC 7.1). The opinion notes that other jurisdictions have treated a chatbot that doesn't identify itself as a chatbot as potentially misleading ².
• No legal advice. Lawyers should make sure the chatbot does not try to give legal advice and should review conversations so mistakes can be corrected ².
• Supervision (RPC 5.3). The lawyer is expected to supervise the vendor, monitor what the chatbot says, and approve autonomous actions such as sending an engagement letter ². Commentary on the opinion highlights that it treats the vendor, not the AI itself, as the party being supervised ⁴.
• Solicitation (RPC 7.3). A chatbot may miss signs that a person is not in a state to choose a lawyer, so the opinion recommends prompt human review and an interview before a relationship goes forward ².

The same commentary notes that the opinion addresses the Rules of Professional Conduct only, not other possible exposure such as contract or consumer protection claims ⁴.

How the two fit together

Read together, the opinions draw a simple line. The 2025 opinion is about lawyers using AI as a tool behind the scenes. The 2026 opinion is about AI standing in front of the public, where the risks shift toward prospective clients, conflicts and the impression the tool gives. Both rest on the same idea: the technology changes, but the lawyer stays responsible for the work and for supervising whoever provides the tool ^{1, 2}.

Where to go from here

• Read both opinions in full on the Oregon State Bar website ^{1, 2}.
• The Oregon Judicial Department has also posted information for lawyers and litigants on using AI tools, after the Oregon Supreme Court issued its first orders over AI-fabricated citations in 2026 ⁵.
• For questions about a specific practice, the Oregon State Bar's ethics resources and a professional responsibility lawyer are the right places to start.

Not legal advice. This post is general information about law and technology, not legal advice. It does not create an attorney-client relationship, and TacticalMS is not a law firm. Laws, rules and ethics opinions change; check the cited sources and talk with a licensed Oregon attorney about any specific situation. Researched and drafted with AI tools from the sources above and checked against our editorial standards. Spot an error? Email intake@tacticalms.dev and we'll fix it.