Texas State Bar AI Ethics Opinion 2024: What Attorneys Must Know

Record Overview

Background and Context

The State Bar of Texas Professional Ethics Committee issues formal opinions interpreting the Texas Disciplinary Rules of Professional Conduct (DRPC). These opinions do not carry the force of disciplinary rules themselves, but they represent the Committee's authoritative reading of how existing rules apply to emerging practice scenarios — and bar disciplinary panels routinely consult them.

By 2024, generative AI tools — including large language model-based drafting assistants, legal research platforms, and document review products — had moved from novelty to routine use at Texas law firms of all sizes. The Committee's 2024 guidance responds directly to that adoption curve, applying longstanding professional responsibility obligations to the specific failure modes that AI tools introduce: hallucinated citations, confidential data exposure through third-party model providers, inadequate supervision of AI-generated work product, and candor questions when AI-assisted filings contain errors.

Core Obligations Addressed

Competence (Rule 1.01)

Texas DRPC Rule 1.01 requires that attorneys provide competent representation, which includes the legal knowledge, skill, thoroughness, and preparation reasonably necessary for the representation. The Committee's 2024 guidance makes clear that this duty extends to understanding the capabilities and limitations of AI tools an attorney deploys in client matters.

Practically, this means an attorney cannot delegate research or drafting to an AI tool and submit the output without independent verification. The competence obligation does not require attorneys to become AI engineers, but it does require enough functional understanding to recognize when AI output is unreliable — including the ability to identify hallucinated case citations before they reach a filing.

Confidentiality (Rule 1.05)

Rule 1.05 prohibits disclosure of confidential client information without consent. The 2024 guidance addresses a specific AI-use scenario: when an attorney inputs client documents or facts into a cloud-based AI tool, the data may be transmitted to and processed by third-party model providers, potentially stored, and in some configurations used for model training.

The Committee's position is that attorneys must evaluate the data-handling practices of any AI tool before submitting confidential client information. This includes reviewing vendor terms of service, confirming whether zero-data-retention or enterprise privacy agreements are in place, and — where the risk cannot be adequately managed — obtaining informed client consent or avoiding the tool for that matter.

Supervision of Nonlawyer Assistance (Rules 5.01 and 5.03)

Rules 5.01 and 5.03 address supervisory obligations over subordinate attorneys and nonlawyer assistants respectively. The 2024 guidance treats AI tools as functionally analogous to nonlawyer assistants for supervision purposes — the supervising attorney retains responsibility for the work product regardless of whether it was generated by a paralegal or a language model.

This framing has direct operational implications. A partner who instructs an associate to use an AI drafting tool is responsible for ensuring that adequate review procedures exist. Firm-level policies that define how AI output is reviewed before use in client matters are not just good practice — under this reading, they are part of the supervisory obligation.

Candor Toward the Tribunal (Rule 3.03)

Rule 3.03 prohibits making false statements of fact or law to a tribunal and requires correction of material false statements. The guidance addresses the scenario — documented in multiple federal court sanctions since 2023 — where AI-generated briefs contain fabricated case citations that attorneys file without verification.

The Committee's position is that citation hallucinations filed with a court constitute a Rule 3.03 violation when the attorney failed to verify the citations before submission. The AI tool's error does not transfer responsibility away from the attorney. The duty of verification is not diminished by the use of AI assistance.

Rule-by-Rule Obligation Summary

Scope and Limitations of This Guidance

The 2024 guidance does not prohibit attorney use of AI tools. It applies existing professional responsibility obligations to AI-assisted practice rather than creating new categories of prohibited conduct. Texas has not, as of the date of this record, enacted a mandatory disclosure requirement specifically requiring attorneys to inform clients or courts that AI was used in preparing work product — though the candor and competence rules may require disclosure in specific circumstances where AI-generated errors are at issue.

The guidance also does not address fee arrangements involving AI efficiency gains, billing for AI-assisted work, or unauthorized practice of law questions arising from AI tools marketed directly to consumers. Those remain open questions under Texas ethics rules as of mid-2026.

Relationship to ABA and Other State Bar Guidance

The Texas guidance is consistent in structure with the wave of state bar AI ethics opinions issued across the US from 2023 onward — California, Florida, New York, and others have addressed similar questions under their own rule frameworks. The ABA issued Formal Opinion 512 in 2024, which addresses generative AI use under the Model Rules, and while Texas does not adopt ABA opinions as binding authority, the substantive positions on competence, confidentiality, and supervision are broadly aligned.

Attorneys admitted in multiple jurisdictions should verify each state's specific guidance rather than assuming uniformity. Disclosure requirements, in particular, vary: some state bars have moved toward explicit client notification requirements that Texas has not yet adopted.

Primary Source

The authoritative text of the Texas Professional Ethics Committee's opinions is maintained at the State Bar of Texas ethics opinions database. Readers should consult that source directly for the full opinion text, any subsequent amendments, and related Committee guidance.