Ethics of AI Technology in Law Practices

Recently, there was an item in the newspaper about a New York lawyer under threat of court sanctions for violating ethical rules. The crux of the matter is that he used a large language model to prepare a legal brief and failed to ensure the sources he cited were accurate. Unfortunately, none of the references provided by ChatGPT were real!

Connecticut attorneys’ ethics rules, like New York’s, do not specifically address using ChatGPT or other AI language models in a law practice setting. The application of AI in a law office is bounded only by the imagination. Client interaction, case management, office administration, motion practice, trial preparation, transactional work, and marketing are only some of the primary categories. The sub-categories are infinite. The Connecticut Rules of Professional Conduct, like the Model Rules of Professional Conduct adopted in many jurisdictions, provides
general guidance on attorney competence, diligence, and the duty to maintain client confidentiality.

While the rules do not explicitly mention AI language models or ChatGPT, attorneys should exercise competence and diligence in their practices. That means ensuring the accuracy, reliability, and relevancy of information. Attorneys should also exercise independent professional judgment when using AI tools and verify the results generated by these systems. Admittedly, AI technology is a new arena for the legal profession just as much as other professional practices and privately held businesses.

Ethical considerations will arise when relying solely on AI language models for research and investigation without conducting additional analysis, verification, and critical assessment of the information. For example, OpenAI’s ChatGPT product utilizes data up to September 2021. ChatGPT is not connected to the Internet. It is not up-to-date. In a profession where laws are interpreted daily by thousands of judges in the United States, a stale date of September 2021 renders ChatGPT useless as a reliable source. Attorneys must provide competent representation to their clients and exercise their professional judgment when using any research tool, including AI technology. As such, AI like ChatGPT and other large language models are not, in my opinion, reliable sources.

Ethical standards evolve, and the appropriate Connecticut authority, the Judges of the Superior Court, may issue additional rules, guidance or opinions addressing the application of AI technology in law practices. Connecticut attorneys should stay informed about any updates or guidance from the Connecticut Judicial Branch, the Connecticut Statewide Grievance Committee, or the Connecticut Bar Association regarding the usage of AI tools in legal practice. Further, non-legal business owners should also adopt policies about utilizing AI in their enterprises.

Consulting with Attorney Bradford J. Sullivan about ethics rules and policies can provide further
clarity on the specific requirements and considerations related to AI-assisted businesses.

Contact Us
​If you are looking for a small firm with big firm experience, look no further than the Law Offices of Bradford J. Sullivan.