Note: The information below is not intended as legal advice.
Artificial intelligence (AI) could make you a more efficient lawyer, but it’s not designed to make you a better lawyer. While AI is already embedded in legal workflows such as research, drafting, and summarization, it cannot and must not replace your professional judgment.
Smart attorneys know that AI is both an opportunity and a liability. Choosing to use AI in your legal practice only heightens the ethical responsibilities you have to your clients and the legal system.
Courts and bar associations across the country are responding to this reality, issuing rulings and guidance to define what responsible AI use looks like in the legal profession. The message is consistent: the technology is moving fast, and professional accountability must keep pace.
Here’s a closer look at what’s driving this moment in AI adoption and what you need to know to stay ahead of it.
The Core Ethical Risks of AI in Legal Practice
The ABA’s Rules of Professional Conduct don’t have an AI carve-out. The same duties that have always governed your practice apply just as squarely when the work is AI-assisted. Three in particular that deserve your attention are: [1]
- Competence. You have a duty to develop and maintain the knowledge and skills to represent your client. No AI tool has ever gone to law school and therefore does not have the specialized expertise required to give credible legal advice.
- Confidentiality. You have a duty to safeguard confidential client information and prevent unauthorized access to it. Using certain AI models could make that sensitive data available to third parties.
- Candor towards the tribunal. You must not knowingly make false statements to the court or fail to correct false statements. Since generative AI is prone to fabricating citations and other facts (known as hallucinating), the onus is on you to ensure its output is accurate. Blind reliance is a breach of your professional duty.
And there’s a fourth consideration worth naming: bias. An AI tool’s outputs are only as good as the data it was trained on. To avoid reinforcing systemic bias, confirm the model you’re using was trained on unbiased and ethically sourced data.[2]
Failure to uphold these duties could result in steep fines and severe reputational damage.
Consider what happened to a seasoned New York attorney who wasn’t aware that ChatGPT could hallucinate, and so never reviewed its output. He submitted a brief containing six citations fabricated entirely by AI. When opposing counsel challenged them, he initially insisted the cases were real. He ultimately paid a $5,000 fine, but the more lasting cost was his credibility.[3]
Leading the Way: How States Are Starting to Define Ethical AI Use
As AI becomes more widely adopted within the legal profession (and as more attorneys make critical errors using it), a growing number of states are issuing guidance on how to leverage it ethically.
Bar associations in California, Florida, New York, Pennsylvania, and Texas have all recently published ethics opinions that, while differing in format, align on core ethical principles, signaling the gradual emergence of a national standard for AI use in legal practice.
The through-line of all these opinions: While AI may take some manual tasks off lawyers’ plates, it heaps on the responsibility to apply the technology with care.
Competence and Accountability Are Foundational
Before you can ethically use AI in your practice, you need to understand how it actually works.
The Professional Ethics Committee for the State Bar of Texas put it plainly in Opinion No. 75: [4]
The lawyer must have a reasonable and current understanding of the technology—because only then can the lawyer evaluate the associated risks of hallucinations or inaccurate answers, the limitations that may be imposed by the model’s use of incomplete or inaccurate data, and the potential for exposing client confidential information.
Learning about AI isn’t something you can check off your list and then forget about. Technological competence is an ongoing obligation that should be part of your long-term professional development plan.
Beyond understanding what AI can and can’t do, you must also understand your role in the legal process. AI is a tool that may streamline your workflow when used properly. However, it’s not the legal expert. You are.
Florida Bar Ethics Opinion 24-1 makes this explicit: [5]
A lawyer may not delegate to generative AI any act that could constitute the practice of law, such as the negotiation of claims or any other function that requires a lawyer’s personal judgment and participation.
Verification and Oversight Are Non-Negotiable
While using AI might help you overcome a blank page, you can’t rely on what it generates to be factually correct. Your eyes must verify every statement the tool makes.
The State Bar of California Standing Committee on Professional Responsibility and Conduct is direct on this point: [6]
[AI-generated outputs] should be critically analyzed for accuracy and bias, supplemented, and improved, if necessary. A lawyer must critically review, validate, and correct both the input and the output of generative AI to ensure the content accurately reflects and supports the interests and priorities of the client in the matter at hand, including as part of advocacy for the client.
Now that AI misuse is on their radar, courts and regulators are on high alert for fabricated citations and other inaccuracies. A missed hallucination can expose you to multiple ethical violations at once (failures of competence, candor to the tribunal, or truthfulness) and trigger sanctions.[5] This is not a theoretical risk. It is happening to attorneys right now.
Confidentiality and Client Protection Take Priority
It’s your duty to safeguard your client’s confidential information. Entering confidential client data into a generative AI tool could put your client’s privacy at risk, especially if you use a free, public version.[7]
The State Bar of California states:[6]
A lawyer should review the Terms of Use or other information to determine how the product utilizes inputs. A lawyer who intends to use confidential information in a generative AI product should ensure that the provider does not share inputted information with third parties or utilize the information for its own use in any manner, including to train or improve its product.
With that in mind, here are a few best practices to guide your approach:
- Avoid putting confidential client information into a tool that can’t demonstrate sound data security practices.[6] Look for SOC 2 certification.[7]
- Anonymize what you share. Avoid entering anything that could identify your client and share only the information necessary to produce the results you need.[3],[6]
- Consider using AI tools that are Enterprise or Legal-Grade and that meet industry encryption standards and ensure data isolation.[7] You may be able to further reduce your risk by developing and using in-house generative AI tools.[5]
- Always obtain consent from your client before using their confidential information in any AI tool for the purposes of working on their case.[3]
- Ask each client in advance whether they have any AI use restrictions. Some corporations, governments, or businesses in highly regulated industries may not permit the use of these tools.[6]
Transparency, Billing, and Client Communication Are Evolving
Disclosing AI use to your clients and the court is the ethical thing to do, and soon, it may be required in the jurisdiction you practice.[2] For instance, two Florida circuits recently issued orders requiring attorneys to disclose their use of AI-generated court filings and to certify their accuracy.[8]
In addition, lawyers in New York City must inform their clients in advance that their discussions will be recorded and summarized using an AI tool. The NYC Bar has noted that people naturally speak differently when they know a verbatim record is being created, and that some may not wish to speak at all under those conditions.[9]
You should also exercise caution when using chatbots on your website. While they can answer basic customer service-oriented questions, they must not dispense legal advice. The chatbot should offer the disclaimer that it’s not an attorney and that using it doesn’t establish an attorney-client relationship.[5]
Transparency is also critical when it comes to billing. You can pass along the actual cost of the AI tool to your client if permitted by law and included in your fee agreement.[4]
What Ethical AI Use Looks Like in Practice
Ethical AI use isn’t just about vetting tools. It’s about updating how your entire practice operates. Here are a few principles to build from:
- Avoid entering confidential client data into unsecured tools.
- Verify the accuracy of every AI-generated output you intend to use in your case.
- Disclose AI use to clients and the court when appropriate or legally required.
- Establish clear and comprehensive internal AI usage policies for your firm and keep them updated.
- Continually train everyone at your firm, lawyers and nonlawyers alike, on your policies and procedures.
At USClaims, we’ve taken this kind of intentional approach to AI innovation ourselves. We use a proprietary generative AI tool specifically to avoid inputting client data into publicly available models.
AI isn’t going away, so its ethical use must stay on your radar from this point forward.
The Risk of Inaction: Why Waiting Is the Bigger Liability
Establishing an ethical framework for AI use will give you and your firm a tremendous competitive advantage. When clients are navigating legal trouble, they’re looking for counsel they can trust. Being transparent about how and when you use AI reinforces your credibility and gives prospective clients confidence in hiring you.
Implementing responsible AI use procedures can also keep you and your firm out of jeopardy. Courts and regulators across the country are actively issuing guidance, and new rules affecting your jurisdiction may already be in effect or arriving soon. It’s far better to build your framework now than to scramble to catch up after a costly mistake.
And the mistakes can be very costly. For instance, a California-based attorney was fined $10,000 for submitting an appeal with 21 fabricated quotes. The lawyer admitted that he didn’t review ChatGPT’s output because he didn’t realize the tool would add citations.[10]
Careless AI use can also ruin your professional reputation. Three Alabama-based lawyers working on the same case got sanctioned for submitting a ChatGPT-generated filing with fabricated citations. However, the judge issuing those sanctions didn’t think fines would be an effective deterrent. Instead, all three attorneys:
- Were disqualified from the case.
- Received public reprimand.
- Received referrals to the state bar.
- Were required to provide copies of the sanctions order to clients, opposing counsel (and every lawyer at their firm), and presiding judges in all pending cases.[11]
It’s worth noting that the law firm employing these attorneys didn’t face any sanctions due to the existence of a robust AI use policy before the incident and a demonstrated commitment to ongoing training after the fact.[11]
Looking Ahead: Responsible Innovation Is the Future of Legal Practice
When used properly, generative AI can be a genuine force for good. But innovation and responsibility have to move together. One without the other isn’t progress.
If your AI use demonstrates a lack of care, prospective clients will look elsewhere for counsel.
Attorneys who set the standard for ethical AI use, rather than waiting to react to court orders, will earn something that no tool can generate: trust.
At USClaims, we use the technology to handle manual work more efficiently; never to substitute for professional judgment, and never at the expense of the people we serve.
Your clients are counting on your expertise to help them through what may be the hardest chapter of their lives. No algorithm has your legal training. No tool has your understanding of nuance. And nothing replaces the compassion you bring to the people who trust you with their cases.
Sources
- American Bar Association. “Model Rules of Professional Conduct – Table of Contents.” 2025. https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/model_rules_of_professional_conduct_table_of_contents/
- Siegel, Daniel J. Philadelphia Bar Association. “Joint Formal Ethics Opinion Gives Practical Guidance On Artificial Intelligence.” 8 Aug. 2024. https://philadelphiabar.org/?pg=ThePhiladelphiaLawyerBlog&blAction=showEntry&blogEntry=111283
- Dunphy, K. Spellbook. “NY Lawyer Fined $5,000 for ChatGPT Fake Citations: Lessons to Learn.” 25 Jan. 2026. https://www.spellbook.legal/learn/lawyer-who-used-chatgpt
- The Professional Ethics Committee for the State Bar of Texas. “Opinion No. 705.” Feb. 2025. https://www.texasbar.com/AM/Template.cfm?Template=/CM/ContentDisplay.cfm&ContentID=67504
- Florida Bar Association. “Ethics Opinion 24-1.” 19 Jan. 2024. https://www.floridabar.org/etopinions/opinion-24-1/
- Standing Committee on Professional Responsibility and Conduct. The State Bar of California. “Practical Guidance for the Use of Generative Artificial Intelligence in the Practice of Law.” https://www.calbar.ca.gov/sites/default/files/portals/0/documents/ethics/Generative-AI-Practical-Guidance.pdf
- Dunphy, K. Spellbook. “State Bar Rules on AI Use: What Lawyers Need to Know About AI Compliance.” 10 Mar. 2026. https://www.spellbook.legal/learn/state-bar-rules-ai-use
- Ash, Jim. Florida Bar News. “11th and 17th Circuits Order Disclosure, Certification of AI Use in Court Filings.” 9 Feb. 2026. https://www.floridabar.org/the-florida-bar-news/11th-and-17th-circuits-order-disclosure-certification-of-ai-use-in-court-filings/
- New York City Bar Association. Committee Reports. “Formal Opinion 2025-6: Ethical Issues Affecting Use of AI to Record, Transcribe, and Summarize Conversations with Clients.” 22 Dec. 2025. https://www.nycbar.org/reports/formal-opinion-2025-6-ethical-issues-affecting-use-of-ai-to-record-transcribe-and-summarize-conversations-with-clients/
- Johnson, Khari. Cal Matters. “California issues historic fine over lawyer’s ChatGPT fabrications.” 22 Sep. 2025. https://calmatters.org/economy/technology/2025/09/chatgpt-lawyer-fine-ai-regulation/
- Artigliere, Hon. Ralph. EDRM. “When AI Policies Fail: The AI Sanctions in Johnson vs. Dunn and What They Mean for the Profession.” 31 Jul. 2025. https://edrm.net/2025/07/when-ai-policies-fail-the-ai-sanctions-in-johnson-v-dunn-and-what-they-mean-for-the-profession/
