LexisNexis Executive Warns Attorneys Face Potential License Loss Over AI Misuse
NEW YORK – A top executive at LexisNexis warns that attorneys are at risk of losing their licenses due too errors adn privacy breaches stemming from the use of publicly available artificial intelligence tools.The warning comes as the legal profession rapidly adopts AI, but grapples with understanding its inherent risks.
According to Sean Fitzpatrick, LexisNexis’s vice president of strategic initiatives, uploading confidential client facts into open-source Large Language Models (llms) like ChatGPT can inadvertently make that data publicly accessible. “You know, information that makes the client identifiable,” Fitzpatrick explained. As thes open-source models don’t protect attorney-client privilege, uploaded data could become discoverable. “I think it’s only a matter of time before we do see attorneys losing their license over this,” he said.
The concern centers around the potential for LLMs to “hallucinate” – generating false or misleading information – and the lack of data security in publicly accessible platforms. Fitzpatrick highlighted LexisNexis’s own generative AI tool, Lexis+ AI, as a potential solution. In March, LexisNexis secured a seven-year contract to serve as an information provider to the federal judiciary.
LexisNexis differentiates its product by stating it does not train its LLMs on customer data and encrypts user prompts. The company also utilizes a “walled garden of content,” a closed, proprietary system updated daily, which Fitzpatrick says is “moast equipped” to solve hallucination issues. However, LexisNexis acknowledges that maintaining privilege remains the responsibility of the attorney.
Experts agree that AI use in legal settings carries inherent risks, regardless of the platform. David Emmert, an expert in the field, categorizes AI models into three types: open-access tools like ChatGPT, in-house “small language models,” and “medium language models” like LexisNexis’ offering.
While firms are increasingly developing in-house AI applications for greater control, Emmert notes these are limited by the comparatively smaller datasets they utilize. Medium-sized models like Lexis+ AI offer broader capabilities but still face questions regarding accuracy and data privacy.
“And the question is, can we fully trust them? … One, that they’re not hallucinating, and second, that the data really remains privileged and private,” emmert said. He cautioned against important investment in AI tools still in their early stages of advancement.
Despite the risks, Emmert believes AI holds significant potential for the legal profession. “Personally, I believe that these AI tools are fantastic,” he said. ”They can really help us get more work done at a higher level of quality with substantially lower investment of time.”
However, he stressed the urgent need for comprehensive education on AI for both legal professionals and students.”Starting in academia but continuing in the profession, we need to train every lawyer, every judge, to become masters of artificial intelligence-not in the technical sense, but using it,” Emmert said. “That’s really where the challenge is.”
