The number of bar associations that have issued AI ethics guidance continues to grow, with NJ being the most recent. In its May 2024 report (Report), the NJ Task Force on Artificial Intelligence and the Law made a number of recommendations and findings as detailed below. With this Report, NJ joins the list of other bar associations that have issued AI ethics guidance, including Florida, California, New York, DC as well as the US Patent and Trademark Office. The Report notes that the practice of law is "poised for substantial transformation due to AI," adding that while the full extent of this transformation remains to be seen, attorneys must keep abreast of and adapt to evolving technological landscapes and embrace opportunities for innovation and specialization in emerging AI-related legal domains.
The Task Force included four workgroups, including: i) Artificial Intelligence and Social Justice Concerns; ii) Artificial Intelligence Products and Services; iii) Education and CLE Programming; and iv) Ethics and Regulatory Issues. Each workgroup made findings and recommendations, some of which are provided below (while trying to avoid duplicating what other bar associations have addressed). Additionally, the Report includes some practical tools including guidance on Essential Factors for Selecting AI Products and Formulating an AI Policy in Legal Firms, provides a Sample Artificial Intelligence and Generative Artificial Intelligence Use Policy and Questions for Vendors When Selecting AI Products and Services, links to which are provided below.
The Report covers many of the expected topics with a focus on:
- prioritizing AI education, establishing baseline procedures and guidelines, and collaborating with data privacy, cybersecurity, and AI professionals as needed;
- adopting an AI policy to ensure the responsible integration of AI in legal practice and adherence to ethical and legal standards; and
- the importance of social justice concerns related to the use of AI, including the importance of transparency in AI software algorithms, bias mitigation, and equitable access to AI tools and the need to review legal AI tools for fairness and accessibility, particularly tools designed for individuals from marginalized or vulnerable communities.
Some of the findings and recommendations are set forth below.
General Findings and Recommendations – Some of the NJ Task Force findings include:
- Transparency on Use of AI – The bench, bar and public must be able to easily determine when AI or generative AI is used in a legal matter. It is equally important to understand how to use AI correctly in light of applicable laws and rules.
- Privacy and Security – When developing or implementing AI systems, collaboration with data privacy experts, cybersecurity professionals and/or AI professionals is highly recommended to ensure responsible integration and adherence to ethical and legal standards. As technology evolves, some of the responsibility for protecting sensitive information may need to shift from law firms to their technology providers, potentially enhancing data protection as these providers are often better equipped to manage sophisticated privacy and cybersecurity challenges.
- Equal Access to Justice Via AI Tools – Ongoing focus is required to ensure that all individuals have access to the technology and tools – including AI – that will allow them to understand and pursue available legal rights and remedies. Guidelines and rules for the use of AI by both pro se litigants and attorneys should prioritize and ensure general awareness and appreciation of the limitations and risks of the use of AI and promote equitable access to AI tools and technologies.
- Equitable Deployment of AI – Partnerships and collaborative networks should be established to promote fair, unbiased and ethical deployment of AI as a resource to be available and applied in appropriate civil and criminal matters, with a particular sensitivity toward justice, equality and ethical considerations.
- Ongoing Review – a structured and reliable framework for ongoing dialogue, review and scrutiny of AI use by courts, legal service organizations and the bar must be established and maintained to ensure appropriate access and implementation of AI for all demographics – in particular underserved, marginalized and at-risk communities.
Regarding lawyer use of AI, the Report stresses:
- Legal professionals must understand that education, knowledge and guidance are necessary to operate AI tools safely and ethically in a legal setting.
- When assessing AI tools and services, it is crucial to categorize them according to their intended users and recipients. Tools designed for the public, as opposed to legal professionals, should not be used for tasks considered "the practice of law." [note: see my comments below on this one]
- When evaluating whether a particular AI tool or service is suitable for its intended use, it is essential to identify and document how data, especially client data, is transmitted, used and stored by the AI to ensure its confidentiality.
- All law firms should adopt an organizational AI policy with a risk assessment framework – as in this sample template provided below.
- Consistent with the guidance from other bar associations, the Rules governing the Courts of New Jersey, the New Jersey Rules of Evidence and the Rules of Professional Conduct are sufficiently flexible to address considerations relating to AI.
On the point about categorizing AI tools designed for the public, as opposed to legal professionals, I respectfully disagree, at least in part. I do agree that some public versions of general purpose AI tools do not have the appropriate safeguards and are not suitable for legal use. However, I do not believe that the proposed categorization of intended use is dispositive, at least to date. I have come across AI tools marketed to legal professionals that do not meet the criteria I would accept for legal use (e.g., inputs are not confidential). Also, some tools designed for the public, especially the enterprise versions of those tools, have the appropriate safeguards. It may make more sense to evaluate any tool, regardless of its intended user base, to assess whether it meets all of the criteria for safe, ethical use by an attorney. As AI evolves, more "standard" software will incorporate AI. This is already happening. For example, certain popular word processing programs are incorporating AI. Although these programs may be deemed designed for the public, they may be perfectly fine for lawyers use as well. In my view each tool needs to be independently evaluated and broad categorizations can be risky.
Some of the practical tools included with the Report are the following.
Essential Factors for Selecting AI Products and Formulating an AI Policy in Legal Firms: A Starting Point – this document provides some initial considerations that law firms and/or legal professionals should consider when integrating AI tools into their practices. It is not a comprehensive list and does not replace a comprehensive AI policy. It covers topics such as risks and advantages of AI, selecting appropriate tools, ensuring data protection and training staff.
Sample Artificial Intelligence and Generative Artificial Intelligence Use Policy – this document provides a sample policy for the ethical, responsible, and efficient use of AI technologies within a law firm to ensure compliance with legal and ethical standards, safeguard client confidentiality and uphold the highest level of professional conduct while harnessing the benefits of AI. This is a good starting point for some of the provisions to include, but each firm will want to enhance and customize this template.
Questions for Vendors When Selecting AI Products and Services – this document provides a sample list of some questions that firms may ask of vendors of AI tools to assess the suitability of those tools for law firm use. This list is useful, but in my view many other questions are necessary when conducting this type of vendor diligence.
As bar associations continue to evaluate the ethical issues with lawyers' use of AI, most are concluding that the existing ethical rules are broad enough to cover AI. However, it is important for lawyers to understand how these rules apply to AI. For more information on this topic see Ethical Considerations for Lawyers Using Generative AI in Delivering Legal Services.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.