Navigating the Attorney’s Duty of Competence in the Age of AI
Generative AI has had a profound impact on the legal profession, and its influence will only accelerate as more litigation-specific AI-tools emerge. For the modern legal professional, integrating these tools into the practice of law likely feels inevitable. When the tools powered by AI are capable of expediting many of the tasks associated with litigation, choosing to work without them may result in working at a disadvantage.
In a world where generative AI continues to bring new tools to the practice of law, one question becomes more and more relevant: Do lawyers have an ethical duty to use generative AI?
Rule 1.1 and Technological Competence
In the American Bar Association’s (ABA) Model Rules of Professional Conduct, Rule 1.1 defines competent representation as “the legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation.” This mandate is further articulated in Comment 8, which stipulates that to maintain requisite knowledge and skill, a lawyer should “keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology.”
The ABA’s definition of competence has been adopted by more than 40 U.S. states and jurisdictions, requiring that lawyers maintain an appropriate level of technological ability. As more AI tools develop, and as AI adoption continues to grow across legal teams, “maintaining technological competence” requires an increasingly solid understanding of AI.
Understanding AI
Generative AI has the ability to create new content by training on large datasets and then producing original outputs that resemble those examples, such as new images, text, and video. Most generative AI systems use large language models (LMMs) to algorithmically predict what word, pixel, or sound comes next in a given sequence. Through this process, AI is capable of generating complete, human-like content.
Agentic AI
Agentic AI is a subset of generative AI, moving beyond single output tasks and instead acting as an autonomous agent (or group of agents) to achieve a specific goal. If the LLMs that power generative AI are the brain, agentic AI is the body, capable of reasoning, planning, and using other tools to complete complex, multi-step tasks with minimal human intervention.
GenAI | Agentic AI | |
Function | Generates single-output content (text, images, video) based on specific user-generated prompts. | Able to complete multi-step tasks autonomously while working toward a well-defined role. |
Scope | Best used for one-off, basic tasks such as summarizing and drafting. This tool needs a greater degree of human collaboration, requiring a new prompt at each step. | Can perform complex tasks such as research and analysis. This tool is best used for multifaceted tasks. |
Benefit | Provides instant access to specific content, answers simple questions from well-defined prompts. | Automates complex processes with the help of human supervision and training. |
Example | Summarizing documents, like lengthy depositions, into digestible content that covers the most relevant information. | Taking action to distribute litigation assignments after cross-referencing a list of unfinished tasks with associate workloads. |
The evolution of generative to agentic AI represents a capability to provide both simple assistance in addition to advanced autonomous work. Competent use of these new capabilities can unlock increased efficiency and improved workflows for any legal team.
Opinions and Regulations
While the efficiencies offered by AI are undeniable, integrating these tools into legal practice necessitates a careful evaluation of the inherent risks, such as data privacy concerns and the potential for bias.
To mitigate these hazards, new decisions are emerging at the state-level to provide clear guardrails:
California SB 574: A proposed bill that would require attorneys to verify materials produced with AI, including case citations and other information involved in case filings.
Florida Bar Ethics Opinion 24-1: Finds that lawyers can use AI in the practice of law, but only if they can protect the confidentiality of client information.
Texas Professional Ethics Committee Opinion 705: Clarifies that attorneys can use generative AI by assuming the obligation to understand how AI works.
Utah Artificial Intelligence Policy Act: Prohibits businesses from blaming AI for consumer protection violations while creating a state regulatory sandbox for AI innovation.
These state-level decisions signal a broader trend toward the standardization of AI within the legal profession. While these tools are being adopted for their exceptional ability to streamline complex workflows, their integration requires more than simple technical deployment.
Guidance from Formal Opinion 512
Formal Opinion 512, issued by the ABA Standing Committee on Ethics and Professional Responsibility, offers formal guidance on integrating AI into the practice of law. The opinion outlines how the core Model Rules of Professional Conduct covering competence, confidentiality, communications, and fees each apply to the use of AI-powered tools.
The committee underscores the importance of this guidance, stating: “With the ever-evolving use of technology by lawyers and courts, lawyers must be vigilant in complying with the Rules of Professional Conduct to ensure that lawyers are adhering to their ethical responsibilities and that clients are protected.”
To that end, the opinion highlights several key provisions that attorneys should closely evaluate when engaging in the use of AI tools, including:
Model Rule 1.1, Competence:
As mentioned, professional competence requires the “legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation.” Formal Opinion 512 translates this standard for practical AI use, explaining: “To competently use an AI tool in a client representation, lawyers need not become AI experts. Rather, lawyers must have a reasonable understanding of the capabilities and limitations of the specific AI technology that the lawyer might use.”
Practically speaking, this means attorneys must either develop a solid grasp of the risks and benefits of the AI tools they use in their work or consult qualified experts who can guide them through the technology's specific capabilities and constraints.
Model Rule 1.6, Confidentiality:
When utilizing AI, lawyers must remain vigilant regarding their obligations under Model Rule 1.6, which requires safeguarding all information related to a client's representation, regardless of its source.
Legal professionals should carefully evaluate whether client information is adequately protected when choosing what AI tools to work with. As the opinion explicitly states: “Before lawyers input information relating to the representation of a client into an AI tool, they must evaluate the risks that the information will be disclosed to or accessed by others outside the firm.”
Model Rule 1.4, Communications:
Legal professionals should independently evaluate whether Model Rule 1.4 requires them to inform the client about AI use in a given instance. Disclosure is mandatory if a client explicitly asks about the tools used to conduct their work, or if the engagement agreement or outside counsel guidelines require it.
Of course, there are instances in which proactive communication is required under Model Rule 1.4. As explained, “clients would need to be informed in advance, and to give informed consent” if an attorney intends to upload information relating to the representation into an AI tool.
Model Rule 1.5, Fees:
Under Rule 1.5, a lawyer’s fees and expenses should always be reasonable. The formal opinion notes that “If a lawyer uses an AI tool to draft a pleading and expends 15 minutes to input the relevant information into the AI program, the lawyer may charge for the 15 minutes as well as for the time the lawyer expends to review the resulting draft.”
However, the opinion goes on to clarify that if using an AI tool “...enables a lawyer to complete tasks much more quickly than without the tool, it may be unreasonable under Rule 1.5 for the lawyer to charge the same flat fee when using the AI tool as when not using it.” Because, of course, a fee charged when little to no work has been done is itself unreasonable.
Morgan v. V2X, Inc
A recent decision from the District of Colorado, Morgan v. V2X, Inc., can serve as a practical, real-world example of both AI use and AI governance in the courtroom. The case involved an employment discrimination suit where the plaintiff, representing himself, used AI tools to help manage the case.
The defendant was concerned that confidential data, such as trade secrets and personnel files, was being fed into AI platforms like ChatGPT or Gemini. The primary risk was that if these tools used the plaintiff's uploads for model training, the privileged information could be replicated in future public outputs.
While the plaintiff argued that his choice of AI tool was protected, the court ruled that, if privileged information is uploaded to an AI platform in a way that could compromise confidentiality, the opposing party has the right to know exactly which platforms are being used.
The court issued an amended protective order with three requirements for using AI with confidential discovery:
Training Bans: The AI provider must be contractually prohibited from using inputs to train or improve its models.
Strict Confidentiality: The provider cannot disclose inputs to third parties unless essential for service delivery.
Right to Delete: The user must have the contractual right to remove or delete all confidential information upon request.
Everlaw’s AI Principles
While mainstream, consumer-grade AI platforms are rarely engineered to meet the strict security and privacy needs of practicing law, there are professional-grade systems that are built specifically to navigate the standards inherent to the legal profession.
Furthermore, AI becomes a powerful, collaborative asset when the technology it powers is anchored by clear principles focusing on ethical use and data security. That’s why Everlaw’s AI-integrations are purpose-built around a foundational set of core principles:
Control: The legal professionals who work with Everlaw maintain control over when AI is used, and data uploaded to Everlaw isn’t used to train public models. Everlaw also uses secure third-party LLMs with no-data-retention policies.
Confidence: Everlaw’s AI outputs are built to be immediately verifiable. When using Everlaw in litigation, content like document summaries provide citations that link directly to the original source.
Transparency: Competent use of generative AI relies on clearly understanding when it is, and isn’t, in use. Everlaw’s user design clearly designates all AI-powered tools and content when used within a workflow.
The Future of Competence
The rise of AI has introduced unique risks and considerations to the practice of law, but it has also delivered one of the most powerful toolkits in the history of the legal profession. Although no explicit ethical mandate currently forces an attorney to adopt AI into their workflows, the foundational obligation to provide competent representation remains absolute. In modern litigation, fulfilling that duty effectively relies on a strong, practical understanding of these new and evolving technologies.
