What are an Alabama lawyer's ethical duties when using AI tools like ChatGPT in legal practice?

Plain-English summary

The opinion addresses how the Alabama Rules of Professional Conduct apply when lawyers use generative AI (tools such as ChatGPT, Gemini, and Claude) and agentic AI (systems that autonomously plan and execute multi-step tasks). Its central conclusion is that AI "does not create new ethical duties for lawyers"; instead it "recontextualizes longstanding professional obligations, competence, confidentiality, supervision, candor, reasonableness of fees, and client communication, within a new technological environment." The lawyer remains fully responsible for the work regardless of whether AI tools were used, citing ABA Formal Op. 512 (2024).

On competence (Rule 1.1), the opinion states that a lawyer need not become an AI engineer but must reasonably understand what a tool does, what it does not do, its known limitations, and how its outputs may fail; selecting a tool without evaluating its accuracy or data-retention policies "may itself constitute a failure of competence." On verification (Rules 1.1, 3.1, 3.3), it states that "AI output must never be accepted at face value," analogizing an AI draft to a first-year associate's first draft that requires substantive review. On confidentiality (Rule 1.6), it treats input into most generative tools as potential disclosure to third parties, distinguishes free consumer versions from contractually governed enterprise systems, and notes AI prompts and logs may be discoverable.

The opinion further addresses client communication and disclosure of AI use (Rule 1.4), supervision of AI as a nonlawyer assistant (Rules 5.1 and 5.3), fees and billing (Rule 1.5), candor to courts and proliferating court-specific AI orders (Rule 3.3), unauthorized practice of law when AI gives unsupervised legal advice (Rules 5.5, 5.3, 8.4), and specific agentic-AI concerns (Rules 1.1, 5.1, 5.3). It closes with a quick-reference checklist for AI-assisted work.

In practice

The opinion holds that, on fees, "lawyers may bill only for time actually spent, and not for the amount of time that would have been spent on a task but for the use of AI"; the supervision and review time is billable, and AI subscription costs "should generally be treated as overhead" unless the client agrees otherwise. It states that submitting AI-generated work without independent verification "can violate, at a minimum, Rules 1.1 and 3.3," with the duty most acute when filing pleadings, citing authority, or making representations to courts.

For supervision, the opinion frames AI as "best understood ethically as a non-lawyer assistant," so lawyers with managerial authority must ensure AI use by others complies with the rules and remain responsible for AI errors. On unauthorized practice, it states that allowing an AI tool to provide legal advice directly to clients without attorney supervision "may constitute UPL" and implicate Rules 5.5, 5.3, and 8.4. Each section is accompanied by the opinion's own "Best Practices" lists, which include verifying every citation against an authoritative database, preferring closed enterprise systems, anonymizing client information, and checking local court rules before every filing.

Common questions

Q: Does using AI like ChatGPT create new ethical duties for Alabama lawyers?

A: No. The opinion states that AI "does not create new ethical duties" but "recontextualizes longstanding professional obligations" such as competence, confidentiality, supervision, candor, and reasonable fees within a new technological environment.

Q: Can an Alabama lawyer file a brief with AI-generated case citations without checking them?

A: No. The opinion states that "AI output must never be accepted at face value" and that submitting AI-generated work without independent verification can violate at least Rules 1.1 and 3.3; it directs lawyers to independently verify every AI-generated citation against an authoritative legal database before filing.

Q: Can a lawyer bill the client for the four hours an AI task would have taken if the AI did it in ten minutes?

A: No. The opinion states that lawyers "may bill only for time actually spent," not the time that would have been spent but for the AI; the lawyer may bill for the time genuinely spent reviewing and correcting the AI's output.

Q: Is it safe to enter client information into a free consumer AI tool?

A: The opinion warns that most generative AI platforms "collect, retain, or reuse user inputs unless expressly prohibited by contract," that entering client information may constitute disclosure to third parties, and that enterprise versions typically prohibit training on inputs while free consumer versions generally do not. It directs lawyers to prefer closed enterprise systems and to anonymize or obtain client consent.

Q: Who is responsible when a non-lawyer staff member uses AI and it produces an error?

A: The supervising lawyer. The opinion states that AI is ethically a non-lawyer assistant under Rules 5.1 and 5.3, and "the supervising lawyer remains fully responsible for the final work product regardless of who operated the AI tool."

Background and rules framework

The opinion interprets the Alabama Rules of Professional Conduct, applying Rule 1.1 (competence), Rule 1.4 (client communication), Rule 1.5 (fees), Rule 1.6 (confidentiality), Rules 3.1 and 3.3 (meritorious claims and candor to the tribunal), Rules 5.1 and 5.3 (supervisory responsibilities over lawyers and non-lawyer assistants), Rule 5.5 (unauthorized practice), and Rule 8.4 (misconduct). These mirror the corresponding ABA Model Rules. The opinion repeatedly relies on ABA Formal Op. 512 (2024) for the proposition that lawyers remain fully responsible for AI-assisted work and must reasonably understand the tools they use.

Citations and references

Rules of Professional Conduct:

• Alabama Rule 1.1 / Model Rule 1.1 (competence, including technological understanding)
• Alabama Rule 1.4 / Model Rule 1.4 (client communication)
• Alabama Rule 1.5 / Model Rule 1.5 (reasonable fees and division of fees)
• Alabama Rule 1.6 / Model Rule 1.6 (confidentiality of information)
• Alabama Rules 3.1, 3.3 / Model Rules 3.1, 3.3 (meritorious claims; candor toward the tribunal)
• Alabama Rules 5.1, 5.3 / Model Rules 5.1, 5.3 (supervision of lawyers and non-lawyer assistants)
• Alabama Rule 5.5 / Model Rule 5.5 (unauthorized practice of law)
• Alabama Rule 8.4 / Model Rule 8.4 (misconduct)

Cases:

• Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023), sanctions for filing a brief with fabricated AI-generated citations
• Ibach v. Stewart, No. SC-2024-0678, 2026 Ala. LEXIS 45 (Ala. Apr. 24, 2026)
• Johnson v. Dunn, 792 F. Supp. 3d 1241 (N.D. Ala. 2025)
• Peebles v. Miley, 439 So. 2d 137 (Ala. 1983), factors for a clearly excessive fee (cited in the fee discussion lineage)

Other opinions cited:

• ABA Formal Op. 512 (2024): generative AI tools and a lawyer's existing duties

See also

• ABA Formal Op. 512: Generative AI Tools
• TX Ethics Op. 705: Generative AI in the Practice of Law
• FL Bar Op. 24-1: Generative AI Use by Lawyers

Source

• Landing page: https://www.alabar.org/office-of-general-counsel/formal-opinions/artificial-intelligence-use-best-practices-under-existing-professional-conduct-rules/
• Original PDF: https://www.alabar.org/assets/2026/05/FO-2026-01.pdf

Original opinion text

Reproduced from the official source for research purposes. The linked source is authoritative.

ARTIFICIAL INTELLIGENCE USE:
Best Practices Under Existing Professional Conduct Rules
Alabama State Bar

Why This Guidance Matters: A Cautionary Example

• In Mata v. Avianca, Inc., 678 F.Supp. 3d 443 (S.D.N.Y. 2023), a lawyer submitted a brief citing more than a dozen cases generated by ChatGPT. The cases did not exist. The court imposed sanctions, finding that the lawyer had failed to verify AI-generated citations before filing. Incidents like this have occurred in courts across the country. This guidance exists to help Alabama lawyers harness the efficiency of AI while avoiding these very real professional consequences. See also Ibach v. Stewart, No. SC-2024-0678, 2026 Ala. LEXIS 45, (Ala. Apr. 24, 2026); Johnson v. Dunn, 792 F.Supp.3d 1241 (N.D. Ala. 2025)

I. Introduction

The Alabama State Bar has always encouraged lawyers to learn the pros and cons associated with various technological innovations in the practice of law, including artificial intelligence. The use of artificial intelligence has tremendous potential to expand and improve the legal system in ways that are only now beginning to be understood. The implementation of artificial intelligence into the customs of law practice will be disruptive, but productively transformative. As more law schools adopt curricula that add, and require, artificial intelligence classes, this technology will become less of a trend and evolve into an actual standard of basic competence. The ABA Model Rules of Professional Conduct (as well as Alabama's Rules of Professional Conduct) require lawyers to provide "competent" representation. The comments to the Model Rules specifically include the knowledge and use of technology as a basic tenet of competence.

Artificial intelligence, and in particular generative artificial intelligence and agentic artificial intelligence, has moved from novelty to daily utility in the practice of law. Lawyers now routinely encounter tools capable of drafting pleadings, summarizing depositions, generating legal research memoranda, predicting litigation outcomes, and responding to client inquiries. These tools promise efficiency, cost reduction, and expanded access to legal services. At the same time, they pose substantial ethical risks if misunderstood or misused.

Contrary to popular perception, artificial intelligence does not create new ethical duties for lawyers. Rather, it recontextualizes longstanding professional obligations, competence, confidentiality, supervision, candor, reasonableness of fees, and client communication, within a new technological environment. As the American Bar Association has emphasized, lawyers remain fully responsible for the work on behalf of the client, regardless of whether AI tools were used in its creation. ABA Formal Ethics Op. 512 (2024).

This guidance applies equally to solo practitioners, small firms, and large firms, although the resources and risk profiles of each may differ. Regardless of firm size, the obligations described herein apply to every Alabama lawyer who uses AI tools in connection with client matters.

Glossary of Key Terms

The following terms are used throughout this document. Lawyers unfamiliar with AI technology should review these definitions before proceeding.

Artificial Intelligence (AI): A broad term for computer systems that perform tasks typically requiring human intelligence, such as understanding language, identifying patterns, and making decisions.

Generative AI (GAI): AI systems that produce new content (text, images, code) based on probabilistic modeling of large datasets. Examples: ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic).

Agentic AI (AAI): AI systems that autonomously plan and execute multi-step tasks with minimal human oversight, interacting with software tools and adapting in real-time.

Hallucination: When an AI system generates false information, fabricated case citations, invented facts, or non-existent statutes, presented with confident, convincing fluency.

Prompt: The instruction or question a user types into an AI system. The quality of the prompt heavily influences the quality and accuracy of the AI's output.

Training Data: The large datasets on which an AI model is trained. The model learns patterns from this data, which shapes what it can and cannot do accurately.

Closed / Enterprise System: A paid, contractually governed version of an AI tool (e.g., ChatGPT Enterprise, Claude for Business) that typically prohibits the provider from using your inputs to train its models, unlike free consumer versions.

Prompt Engineering: The practice of crafting AI prompts carefully to improve output quality and reduce errors such as hallucinations.

Discoverable: Information that may be required to be produced in litigation. AI prompts, drafts, and interaction logs may be discoverable in some circumstances.

II. What Lawyers Mean by "Artificial Intelligence" ("AI")

A. Generative Artificial Intelligence ("GAI")

ABA Formal Ethics Op. 512 defines GAI as technology capable of producing new content, text, images, audio, video, or code, based on probabilistic modeling of large datasets rather than verified legal databases. Popular tools built on GAI include ChatGPT (OpenAI), Gemini (Google), Grok AI (xAI) and Claude (Anthropic).

Unlike traditional legal research tools, GAI systems do not exclusively "retrieve" authority; they predict plausible outputs based on statistical patterns. This distinction is critical. As multiple, and embarrassing, incidents around the country have illustrated, GAI may fabricate facts, citations, and authorities, so-called "hallucinations", with convincing fluency.

B. Agentic Artificial Intelligence ("AAI")

AAI refers to autonomous systems that use reasoning and planning to independently execute multi-step tasks, rather than merely generating content like GAI. These agents work with minimal human oversight, interacting with software tools, analyzing data, and adapting to new situations in real-time. See generally Beth Stackpole, Agentic AI, Explained, MIT Sloan School of Management (Feb. 18, 2026).

To illustrate the practical difference: a GAI tool drafts a discovery response when prompted by a lawyer. An AAI agent, by contrast, might autonomously draft the response when it detects that a new file is opened, email it to the client for review, calendar the filing deadline, and update the case management system, all without further attorney input. This autonomous capability is precisely what makes AAI both powerful and ethically hazardous.

GAI vs. AAI at a Glance. Generative AI (GAI) responds to a prompt from the user, produces content (text, images, code), requires human follow-up to act on output, includes examples such as ChatGPT, Gemini, and Claude, and presents a risk of hallucination in output. Agentic AI (AAI) independently plans and executes tasks, takes actions (emails, filings, calendar entries), operates with minimal or no human oversight, includes examples such as AI legal agents and autonomous workflow bots, and presents a risk of unsupervised errors with real-world consequences.

III. Competence and Technological Understanding (ARPC 1.1)

A. The Standard of Competence Does Not Require Technical Expertise

Under Rule 1.1, a lawyer must provide competent representation, including the "legal knowledge, skill, thoroughness and preparation reasonably necessary" for the representation. Alabama, like most jurisdictions, has adopted the technology-competence principle through commentary rather than new rules.

ABA Formal Ethics Op. 512 clarifies that lawyers need not become AI engineers, but must have a reasonable understanding of: what an AI tool does, what it does not do, its known limitations and risks, and how its outputs may fail.

Vendor selection is also a competence issue. Selecting an AI tool without evaluating its accuracy or data-retention policies may itself constitute a failure of competence. A lawyer who deploys an AI tool without basic due diligence on how that tool works and what it can get wrong has not satisfied Rule 1.1.

B. Ongoing Duty to Stay Informed

Technological competence is not static. Because AI technology evolves rapidly, lawyers must periodically reassess the tools they use, including updates to training data, data-retention practices, and output reliability.

Best Practices: Treat AI tools as you would evolving e-discovery software or cloud platforms, requiring ongoing evaluation, not a one-time review. Incorporate AI literacy into your annual CLE planning. Document your AI use in client matters. Records should include: which tool was used, the date, what the tool was asked to do, what output was generated, and what verification steps the supervising lawyer took. Before deploying any new AI tool, review its published accuracy rates, data-retention policies, and any known limitations or error patterns.

IV. Verification, Judgment, and the Prohibition on Blind Reliance (ARPC 1.1, 3.1, 3.3)

Perhaps the clearest ethical consensus nationwide is this: AI output must never be accepted at face value.

ABA Formal Ethics Op. 512 states unequivocally that submitting AI-generated work without independent verification can violate, at a minimum, Rules 1.1 and 3.3. This duty is especially acute when: filing pleadings or briefs, citing legal authority, and making representations to courts or third parties.

An AI draft should be treated no differently than a first draft from a first-year associate, it requires substantive review, not just proofreading. The lawyer reviewing AI-generated work must bring independent legal judgment to that review, not simply confirm that the output "looks right."

A Note on Prompt Engineering. The quality of AI output depends heavily on how a lawyer constructs the prompt. Vague or poorly crafted prompts increase the risk of hallucinations and off-point output. Lawyers should learn the basics of effective prompting, providing sufficient context, specifying the jurisdiction and applicable rules, and asking the AI to identify its sources. Better prompts produce better, and safer, results.

Best Practices: Independently verify every legal citation generated by AI against Westlaw, Lexis, or a comparable authoritative legal database before filing. Run AI-generated factual summaries against the underlying source documents. Have a second lawyer or senior paralegal review AI-assisted briefs and pleadings before submission. Treat AI drafts as starting points, not finished work product. Never delegate legal judgment, negotiation, or strategic decision-making to AI. Invest time in learning effective prompt construction to improve AI output quality and reduce error risk.

V. Confidentiality and Client Data (ARPC 1.6)

A. Input Is Disclosure

Most GAI platforms collect, retain, or reuse user inputs unless expressly prohibited by contract. Multiple bar opinions and legal authorities across the country have warned that entering client information into AI systems may create privilege risks that courts have not yet uniformly resolved, and may constitute disclosure to third parties under some platforms' terms of service.

Free vs. Enterprise Versions. Enterprise versions of AI tools, such as ChatGPT Enterprise or Claude for Business, typically include contractual prohibitions on using your inputs to train the provider's models. Free consumer versions of these same tools generally do not include such protections. Lawyers should use enterprise or closed systems whenever possible and should understand that the free version of a tool is not the same as the paid enterprise version.

Discoverability. AI prompts, drafts, and interaction logs may be discoverable in litigation in some circumstances, for example, if a party's use of AI to generate a document is placed at issue, or if interaction logs are stored on a system subject to a litigation hold. Lawyers should establish document-retention policies that address AI interactions.

B. Reasonable Safeguards and Informed Consent

Under Rule 1.6, Alabama lawyers must: understand the AI provider's data-handling practices, avoid entering identifiable client information unless adequate safeguards exist, obtain informed client consent when confidentiality risks cannot be eliminated, and understand that AI prompts, drafts, or generated work product may become discoverable in some circumstances.

Best Practices: Use anonymization or substitute identifiers whenever possible when working with AI tools on client matters. Prefer closed, non-training enterprise systems over free consumer AI tools. Review and understand the data-handling and retention terms of any AI tool before using it with client information. Document client consent when disclosure risks remain. Establish firm-wide document-retention policies that address AI interaction logs and AI-generated drafts. AI-generated summaries should not replace preservation of original source documents.

VI. Client Communication and Disclosure of AI Use (ARPC 1.4)

As AI use becomes more prevalent in law practice, clients increasingly expect, and deserve, transparency about how their matters are being handled. While no Alabama rule currently mandates disclosure of AI use in all circumstances, proactive communication is both ethically prudent and consistent with the lawyer's duty of candor to clients.

Lawyers should consider whether to inform clients when AI tools are being used in a meaningful way on their matters, particularly when: AI is used to draft documents that will be submitted to courts or opposing parties on the client's behalf; client-specific information is being entered into an AI system; or AI is used in a way that may affect the fee charged.

Best Practices: Consider including a brief AI use disclosure in your engagement letter, explaining how and when AI tools may be used in the representation. Obtain client agreement on AI billing procedures upfront to establish a transparent and equitable baseline. Respond honestly if a client asks whether AI was used in preparing their documents. Do not mislead clients into believing AI-generated work was entirely human-authored if they ask.

VII. Supervision and Firm-Wide Responsibility (ARPC 5.1 & 5.3)

AI is best understood ethically as a non-lawyer assistant. Much the way a lawyer supervises paralegals and law clerks, lawyers with managerial authority must ensure that AI use by other lawyers and staff complies with professional obligations. A lawyer may be held accountable for errors in AI-generated work product, just as a lawyer is accountable for the errors of supervised staff.

Delegation Risk. A particular concern arises when a lawyer delegates AI use entirely to a non-lawyer staff member without reviewing the output. In this scenario, the staff member may lack the legal training to identify hallucinations, misstatements of law, or ethical problems in the AI's output. The supervising lawyer remains fully responsible for the final work product regardless of who operated the AI tool.

Best Practices: Create written AI-use policies that specify which tools are approved, what tasks they may be used for, and what verification steps are required. Incorporate AI training into your hiring and onboarding process, including training on verification of AI output and confidentiality obligations. Establish review protocols for AI-assisted work, never allow non-lawyer staff to submit AI-generated work product without attorney review. Consider designating an AI Compliance Coordinator or point person responsible for vetting tools, maintaining AI use logs, and fielding staff questions about appropriate use. Centralize approval of AI tools used firm-wide. Treat AI governance with the same level of care as other compliance obligations, such as trust account management.

VIII. Fees, Billing, and Reasonableness (ARPC 1.5)

As AI technology creates efficiencies, it will inevitably affect rates and billing structures. Efficiency does not justify inflated billing. Multiple bar opinions caution that lawyers may bill only for time actually spent, and not for the amount of time that would have been spent on a task but for the use of AI.

The supervision and review time is billable. If AI drafts a brief in ten minutes that would have taken four hours, the lawyer cannot bill four hours as if the AI did not exist. However, the lawyer can bill for the time spent reviewing, correcting, and exercising legal judgment over the AI's output, because that review time reflects genuine professional work. This is analogous to the time spent reviewing a junior associate's draft.

AI subscription costs are overhead. AI tool subscription costs should generally be treated as overhead, not billed directly to clients, unless the client has specifically agreed otherwise, similar to how Westlaw or Lexis subscriptions have historically been treated by most firms.

Best Practices: Disclose AI use to clients when it affects the fees charged. Avoid double-billing or padding time saved through automation. Bill only for time actually spent on the matter, including genuine review and judgment time. Treat AI subscription costs as overhead unless otherwise agreed upon with the client in writing. Obtain client agreement on AI billing procedures upfront to establish a fair and transparent baseline.

IX. Candor to Courts and Disclosure Obligations (ARPC 3.3)

Some courts now require disclosure or certification regarding AI use. Even where disclosure is not mandated, lawyers remain responsible for ensuring that AI-assisted filings are accurate and truthful.

Court-Specific Requirements Are Proliferating. Numerous federal district courts and state courts have issued standing orders governing AI use in filings, ranging from mandatory disclosure to certification that AI-generated content has been independently verified. Lawyers should check local rules and standing orders before every filing in every court where they practice. This is not a one-time task, court rules on AI are evolving rapidly.

Best Practices: Before filing in any court, check that court's local rules and any standing orders for AI disclosure or certification requirements. Never allow AI to generate factual or legal assertions that are submitted to a court without independent attorney verification. When a court order requires certification of AI use, ensure that certification is accurate and complete. Treat the duty of candor to tribunals as the floor, not the ceiling, for AI-related disclosure.

X. Unauthorized Practice of Law (ARPC 5.5, 5.3, 8.4)

Because Alabama lawyers increasingly interact with AI-generated legal products, it is important to address when AI use could cross into or facilitate the unauthorized practice of law ("UPL").

Example. A law firm deploys a chatbot on its website that answers prospective clients' legal questions in real-time. If the chatbot provides specific legal advice without flagging that it is not an attorney and without attorney review of its responses, this may constitute UPL, and the lawyer who deployed it may be responsible under Rules 5.5, 5.3, and 8.4.

Key concerns include: Lawyers cannot allow AI tools to independently provide legal advice directly to clients without attorney supervision. Law firms should avoid marketing AI tools in a manner suggesting that the software itself is "the lawyer." Lawyers are responsible for all legal advice and information generated through AI-assisted systems. Outsourcing legal judgment to AI implicates Rule 5.5 (unauthorized practice), Rule 5.3 (supervision of non-lawyer assistants), and Rule 8.4 (misconduct), including if the lawyer knew or should have known the AI was providing unsupervised legal advice.

XI. Specific Agentic AI Concerns (ARPC 1.1, 5.1, 5.3)

The benefits and risks of AAI lie in its capacity to perform multifaceted, autonomous tasks with little or no human oversight. Unlike static processes, AAI systems evolve as they learn, mostly without human interaction or supervision. This autonomy enhances utility and value, but may disproportionately increase the risk of harm to clients and noncompliance with professional conduct rules. Rule 1.1 requires that lawyers understand the benefits and risks associated with AAI before deploying it in practice.

Some experts have argued that embedding "core values" or governing commandments into AAI systems when created can produce more reliable results. One author has proposed the following framework for legal agentic AI:

Proposed Commandments for Legal Agentic AI (Steven M. Cordero, 10 Commandments for Agentic AI Tools in the Legal Industry, Akerman.com (Dec. 5, 2025)):

I. Legal agentic AI shall only accept prompts from the attorney.
II. Legal agentic AI shall adhere to the ABA Model Rules of Professional Conduct or state-law equivalents wherever the attorney is admitted to practice law.
III. Legal agentic AI shall not engage in dishonesty or deception, nor perform tasks that will aid in dishonesty or deception.
IV. Legal agentic AI shall not impersonate a human being.
V. Legal agentic AI shall not access nonpublic information unless expressly directed to do so by the attorney.
VI. If legal agentic AI is provided nonpublic information, it shall only allow access to that information to the attorney.
VII. Legal agentic AI shall not hallucinate, rely on, or create fictitious sources.
VIII. Legal agentic AI shall disclose to the attorney all sources of information relied on in performing its tasks.
IX. Legal agentic AI shall not rely on sources created by other AI unless expressly directed to do so by the attorney.
X. Legal agentic AI shall perform all tasks as instructed by the attorney, act proactively, offer insights, and suggest actions when appropriate, so long as such performance does not conflict with the first through ninth commandments.

As the author of these commandments acknowledges, AI is not so reliable that embedding core values alone will ensure compliance or prevent harm. Attorney oversight will always be critical, regardless of attempts to program AI with governing principles.

Best Practices: Do not deploy AAI tools on client matters without a documented human review checkpoint before any action is taken on the AI's behalf. Define in writing what tasks AAI is and is not authorized to perform autonomously. Audit AAI activity logs regularly to identify errors, unauthorized actions, or outputs that require correction. Treat AAI governance as you would any other high-risk workflow requiring documented protocols and supervisory sign-off. Consult with AI vendors about what safeguards and audit trails are built into their agentic AI products before deploying them.

XII. AI Use Quick-Reference Checklist for Alabama Lawyers

The following checklist is designed for day-to-day use before and during AI-assisted legal work. Before Using AI in a Client Matter, Have You: Confirmed that the AI tool I am using is an approved, enterprise-grade system with data protections, not a free consumer version. Reviewed and understand this tool's data-retention and training policies. Anonymized or removed identifiable client information from my prompts, or obtained client consent for its inclusion. Crafted my prompts carefully to include relevant context, jurisdiction, and applicable rules. Reviewed all AI-generated content myself and applied independent legal judgment, I have not simply accepted the output. Independently verified every case citation, statute, and factual assertion in AI-generated content against authoritative sources. Checked this court's local rules and standing orders for any AI disclosure or certification requirements (if filing). Documented my AI use for this matter, including the tool used, the task performed, and the verification steps taken. Confirmed that no legal judgment, strategic decision-making, or client advice was delegated to the AI. Reviewed any AI-generated client communications before they were sent. Ensured that non-lawyer staff who used AI on this matter did so under adequate attorney supervision.

XIII. Conclusion

Artificial intelligence does not lower ethical standards, it raises the stakes for compliance. For Alabama lawyers, the safest course is not avoidance but informed, disciplined, and transparent use of AI tools. All available ethical and legal authority on these issues makes one principle unmistakable: technology may assist the lawyer, but it never replaces the lawyer.

Used properly, AI can enhance competence, efficiency, and access to justice. Used carelessly, it invites sanctions, disciplinary exposure, and client harm. The ethical path forward lies not in novelty, but in fidelity to foundational professional duties, applied rigorously in a new technological age.

Should you have any questions about your ethical obligations concerning the use of artificial intelligence, please contact the Alabama State Bar. You can reach us by telephone at 334-269-1515 or by emailing [email protected].