In a recent legislative move, California Assemblymember Josh Lowenthal is spearheading efforts to establish clear regulations on the application of artificial intelligence (AI) by legal professionals, particularly concerning the autonomous generation of text and content for court filings. The proposed bill, known as A.B. 2811, aims to introduce disclosure and citation requirements for AI-assisted legal documents. While specific details of these requirements remain under development, Lowenthal’s office is diligently crafting precise language to address this emerging issue.
The initiative reflects growing concern within the legal community over the implications of AI technologies, such as ChatGPT, which have already led to notable incidents, including the submission of briefs citing fictitious cases. Bradford Hise, a legal ethics advisor, emphasized the rapid evolution of this field and the various challenges it presents for legal practitioners.
Many lawyers already leverage AI for tasks like research and brief analysis, utilizing tools provided by platforms like Westlaw and Bloomberg Law. However, the use of generative AI technologies, capable of entirely producing legal documents, raises significant accuracy and reliability concerns. Hise highlighted the importance of ensuring the precision of all aspects of AI-generated work, from text to legal conclusions, to mitigate risks of misinformation.
At the national level, some policymakers have initiated measures to guard against the adverse effects of inaccurate AI on the legal profession and its clients. For example, a policy memorandum from a New York City borough’s office has prohibited the use of generative AI for providing legal advice, citing the technology’s lack of current legal understanding.
In response to the technology’s proliferation, the California State Bar issued guidelines last November, advocating for transparency in AI usage among lawyers and rigorous human review of AI outputs. These measures aim to prevent overreliance on AI and ensure the identification and correction of any inaccuracies or biases.
Lowenthal’s bill, aligning with these guidelines, seeks to enhance disclosure practices and citation accuracy, urging collaboration between the state bar and legislature on potential regulatory changes for legal AI products.
Despite these efforts, some legal analysts question the necessity of such legislation, pointing out existing mechanisms that penalize the submission of false or inaccurate information. Critics argue that the legal profession should regulate itself, considering the rapid pace of AI advancements and the potential for current technologies to become obsolete.
This legislation underscores the complex interplay between innovation and accountability in the legal sector, highlighting the need for adaptable frameworks that can evolve alongside technological advancements.
