Why ChatGPT Can’t Be Your Attorney

Thinking about using AI to represent yourself in litigation? Think again. It is well-documented that pro se litigants (those who represent themselves) are less likely to prevail compared to their represented counterparts, due to a lack of procedural and substantive legal knowledge. Regardless, AI has provided people with a false sense of security, tempting them to represent themselves regardless of the statistical realities and consequences. For experienced attorneys – like those at Lavelle Law – AI errors are obvious. But for pro se litigants, blind confidence in AI-generated “legal advice” can be costly.

More judges are becoming less tolerant of litigants making misrepresentations in pleadings based on what AI tells them. As the use of AI in the courtroom has increased, courts have not hesitated to impose sanctions for AI misuse where parties cite fake cases and false arguments due to overreliance on AI. Sanctions may include monetary penalties, requirements to donate to a legal charity, or dismissal.

Why is it Unreasonable to Rely on AI for Legal Advice?

There has been adequate warning from multiple trusted sources that legal information provided by AI is prone to contain errors and cite inaccurate or nonexistent cases.

In 2024, the American Bar Association issued Formal Opinion 512 regarding AI, in which it warns that generative AI tools are prone to “hallucinations.” Hallucinations in this context have become known in the legal community as AI responses that have no basis in fact or reality. The ABA explains that AI is trained on limited and potentially outdated content, with tools that lack the ability to understand the meaning of text or evaluate context. As a result, AI hallucinations often cite nonexistent cases as fake support for a statement.

Many AI systems are not trained on case law but rather only rely on secondary sources to generate responses that simply sound right – like Reddit comments, for example. A study from Stanford concluded that the statistical probability that an AI response contains factual errors, irrelevant information, or is not grounded in the jurisdictions governing case law is much too high for parties to trust the AI response blindly for legal research. Therefore, AI should never be used to supplant an attorney’s professional judgment.

Hiring a licensed attorney brings the following benefits that AI simply can’t compete with:

1. Representation required by law – most businesses are legally required to be represented by a licensed attorney in court.
2. Courtroom advocacy – reading the judge’s tone and the jury, adapting arguments in real time, and persuasive candor.
3. Novel legal issues – attorneys are equipped to think critically and fill in the gaps in the law with experience in other cases and persuasive arguments based on other jurisdictions.
4. Confidentiality – having a real attorney creates the attorney/client privilege.
5. Relevant arguments – attorneys are trained to identify which facts are important and which are irrelevant.
6. Settlement advantage – negotiation skills require human judgment, persuasion, nuance, and appeal to emotions.

If you’re considering representing yourself in litigation, contact Lavelle Law Attorney Sarah Reusché instead at sreusche@lavellelaw.com or 847-705-7555. Free initial consults are available.