AI, My Co-Counsel: The Use of Artificial Intelligence in the Legal Community

Written by Savanna Burney

Savanna Burney served as the inaugural Student Reporter for the CLA’s 2025 Annual Meeting, September 2025. She is a third-year law student at the University of California, Davis School of Law, with aspirations to become a litigator. Savanna is highly involved in her community, participating in Moot Court, the Negotiations Team, the Humanitarian Aid Legal Organization, and the American Bar Association as Vice Chair of the Law Student Division.

Introduction

In a fast-paced career field, saving a few hours can be the difference between being fully rested and falling asleep in the courtroom. For decades, attorneys have shouldered the burdens of routine but time-consuming tasks such as drafting briefs, preparing motions, and conducting exhaustive legal research. It is only logical that when a tool was developed to streamline these processes, the legal community would begin to adopt it. In the same vein as libraries being condensed to internet searches, drafting can be assisted by large-language models.

Artificial Intelligence “AI” was created as a tool to help its users but at some point, unlimited access to it has presented issues. For example, an attorney relying on AI for case research may find that the model “hallucinates” or fabricates cases; these errors can have serious professional consequences. As such, these hallucinations have initiated a call around the country for regulation and guidance with some wanting to ban the technology for the general public and others suggesting the government to intervene.

At the Annual Meeting of the California Lawyers Association, a panel of individuals in the legal field explored their experiences with AI and what attorneys can do in the face of this changing technology. The panel’s message was clear: while AI is not yet advanced enough to replace attorneys, it is already reshaping the practice of law.

California Lawyers Association – “The Law Will be Automated”

In a session titled “The Law Will Be Automated: Will You?”, a panel that included Hallie Balkin, Jeremy Evans, Tim Sawyer, and Dominique Shelton Leipzig convened to share their insights on the AI technologies and the impact of them in the legal profession.

The panel described three things that an attorney needs to do well: (1) aggregate information to solve the problem, (2) analyze the information received, and (3) plan a course of action. AI can be helpful in any of these tasks as long as it is utilized appropriately. The panel proceeds to describe thinking of AI as a “calculator” or as “taking care of a child or the elderly” with the intent that AI needs to be carefully monitored. The panel also delivered statistics to help users better understand how well the technology works. AI models currently have a 40-47% error rate in their responses. However, a combination of human input and AI can have a 95%+ success rate, or a less than 5% error rate. Furthermore, the panel warns users to check AI “at every turn”. The panel also warns attorneys that law firms that use AI will replace those who do not, which emphasizes its importance in this growing realm of technology. Lastly, the panel ensures that regulations are coming from the government and that measures are already being proposed in other countries such as the United Kingdom.

Opinion

The AI panel from the California Lawyers Association was knowledgeable about the subject and helpful in their assessment of the current AI tool. Attending the panel was valuable to my understanding of AI and how to engage with the technology while making me feel more comfortable in using it. However, after attending the panel, I feel more confident in using the technology in limited aspects such as summarizing key information or helping me write non-critical emails.

The panel was very insightful in their perspective of AI and these insights should be further shared. The panel illustrated the point of AI hallucinations with the following example: if you ask AI what your friend had for breakfast, it will confidently provide an answer, despite having no factual basis. AI has hardly ever admitted it does not know the answer which I believe is a major flaw. Imagine if every time someone asked you a question, you could not say that you did not know. The result would likely be many misadventures as well as detrimental outcomes. As search engines are able to be without a response, so should AI.

Further, the panel rightly discussed how technology induces change, specifically with the example that AI is like a calculator. Just as people believed that we would never have a calculator in our pocket, and as such would need to learn mathematics, AI is akin to this idea. Technology has evolved to such a point that we are able to have calculators in our pockets, as well as a phone to call, the Internet, and games. The calculator comparison reminded me of a recent conversation that I had in a group of friends. Movies in the late 2000s to mid-2010s made jokes on how Siri could not understand people and hilariously, would misinterpret words. These jokes have faded as Siri has improved and no longer misunderstands its user. At some point, AI will become this. Now, we laugh at silly AI videos and pictures that are obviously doctored. However, one day, these photos and videos will likely become so realistic that it will be difficult to distinguish them from real media. In that vein, concern for AI is warranted and protections for users should be encouraged from our government. However, with concern should come appreciation. Our technology has evolved to such an extent that we are able to better be creative and gain assistance quicker and one day, more reliably.

Ultimately, any tool needs the appropriate supervision to serve its user. As a power drill needs direction, so does an AI technology. It can be used easily for mindless tasks or being an extra pair of eyes on a document. In the scope of research, it is not yet intelligent enough to truly replace hard work and time. Expanding on a principle discussed in the session, learning how to use AI is necessary for attorneys to continue progressing. Eventually, this tool will become fine-tuned to truly be of use to attorneys and help shave off hours of work. It is better to learn sooner how to integrate these technologies rather than catch up later.

Conclusion

After attending the Annual meeting and this panel, I am grateful for the opportunity to learn from such esteemed leaders in the profession. With this new knowledge and confidence, I am able to use AI in a way that will help my career but also be mindful of its limitations. With every tool, there are downsides. The major takeaway from this session is that our career field will keep progressing and if we do not take the time to learn how to be better, we will be left behind. A partnership between a human and an AI leads to a 95+% success rate and it is our duty to be successful for our clients. Lead with passion but lead carefully and with consideration that each choice will affect someone.

Thank you all for reading my article and I sincerely hope that it inspires you to take a chance on AI, with purpose and integrity.