A few months after the introduction of OpenAI’s ChatGPT, some lawyers in New York City employed the AI tool to compose a legal brief. AI erroneously referenced fictitious legal cases. This incident served as a reminder that, even though AI like ChatGPT can be valuable, it is essential to review their work meticulously.
This incident did not go unnoticed by LexisNexis. LexisNexis is a legal software company that provides tools to assist lawyers in locating case law to bolster their legal arguments. LexisNexis acknowledges the potential of AI in streamlining the repetitive aspects of legal work commonly undertaken by lawyers. Nevertheless, the company acknowledges the real challenges as it embarks on its journey into generative AI.
LexisNexis has been integrating AI technology into its platform for a significant period. Incorporating functionality similar to ChatGPT into their legal toolkit could enhance lawyers’ efficiency. It comes in handy in tasks such as brief writing and citation retrieval.
However, Jeff Reihl, LexisNexis Chief Technology Officer, and his team are aware of the associated risks. LexisNexis recently surveyed 1,000 lawyers to gauge their perspective on the potential of generative AI in their profession. The results showed a generally positive sentiment. However, the results underscored lawyers’ awareness of the technology’s limitations, tempering some initial enthusiasm.
A prominent challenge in this context is establishing trust in the information generated by AI bots. The legal software is actively working on multiple strategies to instill confidence in the results provided by their AI tools.
LexisNexis Fixing AI’s Problems
The truth remains that any Language Model (LLM) is bound to produce wrong information. LexisNexis is actively addressing this problem through several measures. They are training their models using their extensive legal dataset to mitigate some of the reliability issues associated with foundational models. Reihl explains that the company also incorporates the most up-to-date case law from their databases.
LexisNexis is taking steps to enable lawyers using the software to understand how the AI arrived at a particular answer by providing a reference to the source.
This whole process is an ongoing development. The legal software is currently collaborating with six clients to refine their approach based on feedback. The plan is to release AI-powered tools in the coming months.
The featured image is from LexisNexis.com
