As personal injury attorneys, you are accustomed to toggling between the BI and PD claims in a case, while managing NPRs and responses to RFPs. If you haven’t done so already, be prepared to incorporate another acronym into your repertoire – AI. But please understand, it does not stand for “accurate information.”

Artificial intelligence is here to stay, and clients expect us to use it and pass along the cost savings associated with its efficiencies. In many personal injury practices where high-volume caseloads are critical to the firm’s bottom line, AI can streamline processes that you perhaps never dreamed would be possible. That stack of medical records on your paralegal’s desk? It can be summarized in mere seconds, producing the timeline you need to detail your client’s care after an accident. Need to draft a comprehensive demand letter to send to the at-fault party’s insurer? An AI model can immediately generate a letter that would have taken your associate hours to develop. However, relying on AI alone without attorney oversight is a legal malpractice claim waiting to happen.

While the use of AI in the legal world will likely become even more widespread this year, any “work product” that an AI model provides must be carefully scrutinized for accuracy. As you may have already learned the hard way, an AI platform may offer a response which appears to be legitimate instead of admitting it doesn’t “know” something. And despite the seemingly daily news articles about lawyers being disciplined or fined for filing briefs or pleadings containing false information, AKA AI hallucinations, practitioners from law firms of all sizes continue to submit documents to courts without properly reviewing them for errors. So, here are some tips FYI as you incorporate AI into your practice:

Keep it on the DL – Unless you have a captive AI platform solely for use within your firm, you must assume that any information you enter into the AI model may be viewed by anyone. In other words, you cannot have any expectation of confidentiality when using these tools. By including a client’s confidential information in your prompt, you are breaching your ethical duty to not reveal any information related to the representation under Rule 1.6. For example, asking ChatGPT about the life expectancy of your client based on their DOB and the extent of the injuries they suffered in an accident would be a clear violation of the Rule. 

Double check the INFO – As noted above, you cannot assume that the information generated by an Al model – including the purported case citation – is correct (or even real). Even bespoke legal AI tools hallucinate a surprising percentage of the time. At the time of publication in May 2024, a Stanford University study showed that Lexis+ AI and Ask Practical Law AI produced inaccurate information in 17% of its responses, with Westlaw’s AI tool hallucinating over 34% of the time. While the technology will continue to improve, these statistics demonstrate how crucial it is for lawyers to Shepardize any cases the AI model uses in its response. Lawyers should also fully evaluate any legal arguments it presents as reasonable. While AI can increase efficiencies in your practice, it is not a substitute for the critical thinking and research skills you have worked hard to develop over many years of practice.

Update your SOPs – If you do not have a firm policy for the use of AI, create one ASAP. Consider how you will permit other lawyers and non-lawyer assistants to use it and for what purposes. You also need to determine how you will communicate to your clients that you will be incorporating AI into your practice and how you will keep their information confidential. Consider consulting this recent ABA Checklist for using AI responsibly in your firm. And because this technology continues to evolve at a dizzying pace, be prepared to revisit and revise your AI policies and procedures frequently.  

Whether you are fully comfortable exploring how AI can benefit you and your clients or are treading lightly when it comes to artificial intelligence, at the EOD we all have an obligation to stay abreast of the risks and benefits of technology under Rule 1.1 and to learn how to ethically use it in practice.