AI errors trouble courts worldwide
New Delhi | Desk: A growing number of errors in legal documents prepared using artificial intelligence (AI) abroad have posed a new problem for judges around the world.
According to lawyers and court documents, AI-generated briefs often include citations to cases that do not actually exist.
This trend has come as a warning to those using AI tools in the workplace. From legal to accounting and marketing, many professionals are using AI chatbots and assistants to increase productivity, but they are also finding that these programs can make mistakes.
French data scientist and lawyer Damien Charlotin has cataloged at least 490 court filings in the past six months that contained false or misleading information, including "hallucinations" generated by AI. He says this pace is accelerating as the use of AI increases. Charlottesville has created a database tracking cases where judges found that AI fabricated case law and provided false citations.
The severity of the errors is such that a federal judge found nearly 30 faulty citations in a brief filed by a lawyer for MyPillow Inc.
Workplace experts say AI should be viewed not as a coworker (intern), but as an assistant that improves your workflow. Maria Flynn, CEO of Jobs for the Future, warns, "Key tips to avoid the risk of trusting AI to make major decisions for you:
Check accuracy: According to Atlanta attorney Justin Daniels, "People assume it's accurate because it seems so reliable." Take the time to verify the accuracy of factual information generated by AI, even if it's inconvenient and time-consuming.
Consider privacy: When using free AI tools, don't enter identifying information or corporate secrets. Once uploaded, it's possible that other people using the same tool could also access that information, as AI doesn't distinguish between public and private data.
Seek legal advice: Before using AI to record meetings, consider whether the conversation should be kept privileged and confidential, and consult with legal or HR departments.
Experts say that despite these potential problems, developing AI fluency is extremely important. In Flynn's words, "Not learning to use AI is the only way to avoid it." The biggest possible mistake in learning this is "Don't believe it."
