State Bar AI Ethics Opinions: A Comprehensive Guide (2026)

As AI tools like ChatGPT and Claude become embedded in legal practice, state bars are scrambling to provide guidance. The ABA Model Rules offer a framework, but specific guidance varies by jurisdiction — and it's evolving rapidly.

This guide compiles AI ethics guidance from state bars across the country. We'll update it as new opinions are issued.

Summary Table

California

California has been a leader in AI ethics guidance. The State Bar of California released "Practical Guidance for the Use of Generative Artificial Intelligence in the Practice of Law" in November 2023.

Key Points

• Competence (Rule 1.1): Attorneys must understand AI capabilities and limitations before using these tools. This includes knowing whether prompts are used for training.
• Confidentiality (Rule 1.6): Attorneys must evaluate whether inputting client data into AI tools could result in unauthorized disclosure. Consumer-grade tools may present unacceptable risks.
• Supervision (Rule 5.3): Work product generated by AI must be reviewed "with the same rigor as work from any other source."
• Candor: Attorneys should not misrepresent AI-generated content as their own work when that would be misleading.
• Billing: Time spent reviewing and correcting AI outputs is billable; time saved through AI efficiency may need to be reflected in fees.

Practical Recommendations

California's guidance suggests attorneys:

• Review AI tool terms of service and privacy policies
• Consider using enterprise versions with stronger privacy protections
• Avoid inputting confidential client information when possible
• Verify all AI outputs, especially legal citations
• Disclose AI usage to clients when material

Florida

The Florida Bar's Professional Ethics Committee issued Proposed Advisory Opinion 24-1 addressing AI usage in legal practice.

Key Points

• Confidentiality is paramount: Attorneys must ensure client information entered into AI tools remains protected. The opinion notes that data entered into AI platforms may be retained, used for training, or disclosed.
• No blind reliance: AI outputs must be verified. The opinion specifically cites the Mata v. Avianca case as a cautionary example.
• Disclosure requirements: Attorneys may need to disclose AI usage to clients, particularly when AI significantly contributes to deliverables.
• Fee considerations: Charging the same rate for AI-assisted work as fully human work may raise fee-related ethics issues in some circumstances.

New York

The New York State Bar Association (NYSBA) released a comprehensive report from its Task Force on Artificial Intelligence in April 2024.

Key Points

• Technology competence: Aligns with ABA Model Rule 1.1's technology competence requirement. Attorneys must understand AI tools before using them.
• Confidentiality concerns: The report emphasizes that attorneys must understand how AI providers handle data and whether inputs are used for model training.
• Supervision obligation: AI should be treated as a tool requiring supervision, similar to delegating work to a junior associate.
• Firm policies: The NYSBA recommends that firms adopt written AI usage policies.

Court Rules

Several New York courts have also implemented AI-specific rules requiring disclosure of AI use in court filings, including certification that AI-generated content has been verified.

Texas

The State Bar of Texas established an AI task force that has issued recommendations for attorneys using AI tools.

Key Points

• Competence includes AI literacy: Attorneys should understand the basics of how generative AI works, including its tendency to "hallucinate."
• Confidentiality review: Before using any AI tool for client matters, attorneys should review the tool's privacy policy and terms of service.
• Human review required: All AI outputs must be reviewed by a competent attorney before use.
• Documentation: Consider documenting AI usage for client files.

New Jersey

New Jersey's Office of Attorney Ethics issued guidance on AI in legal practice.

Key Points

• RPC 1.6 compliance: Inputting confidential client information into AI tools may violate confidentiality rules unless proper safeguards are in place.
• Enterprise tools preferred: The guidance notes that enterprise-grade AI tools with appropriate data handling may be more appropriate for legal work than consumer versions.
• Verification required: Attorneys must verify AI outputs and cannot rely on AI-generated citations without independent verification.

Colorado

The Colorado Bar Association issued Formal Ethics Opinion 145 addressing AI.

Key Points

• Confidentiality first: Attorneys must ensure that using AI tools does not result in disclosure of confidential information.
• Reasonable measures: Attorneys should take "reasonable measures" to protect client data when using AI, which may include using enterprise tools, avoiding sensitive data input, or anonymizing information.
• Ongoing duty: The duty to understand AI tools is ongoing — attorneys must stay current as tools evolve.

Common Themes Across Jurisdictions

While specific language varies, state bar guidance converges on several key principles:

1. Attorney Responsibility Remains

AI is a tool, not a replacement for attorney judgment. Attorneys remain fully responsible for work product, regardless of how it was generated.

2. Confidentiality is Non-Negotiable

Every state emphasizes that client confidentiality obligations apply to AI usage. Attorneys must understand how AI tools handle data before inputting client information.

3. Verification is Required

AI outputs — especially legal citations — must be independently verified. The Mata v. Avianca case serves as a universal cautionary tale.

4. Policies Should Be Written

Multiple state bars recommend that firms adopt written AI usage policies defining acceptable use, review requirements, and confidentiality protocols.

5. Disclosure May Be Required

When AI significantly contributes to client deliverables, disclosure to clients may be appropriate or required.

What This Means for Your Firm

Regardless of your jurisdiction, the path forward is clear:

1. Understand your tools: Know how your AI platforms handle data, whether prompts are used for training, and what the provider's confidentiality commitments are.
2. Implement safeguards: Use enterprise-grade tools, implement technical controls, and train your team on responsible AI use.
3. Document everything: Maintain records of your AI policies, training, and safeguards in case questions arise.
4. Stay current: AI guidance is evolving rapidly. Monitor your state bar for updates.

This article is for informational purposes only and does not constitute legal advice. For specific guidance on AI ethics compliance in your jurisdiction, consult your state bar or a legal ethics professional.