Raghu Ramanathan, president of Legal Professionals at Thomson Reuters, discusses the evolution and future of generative artificial intelligence (GenAI) in the legal sector, highlighting a shift from rapid model development to improving efficiency and delivering better outcomes. It emphasises the need for meaningful benchmarks that reflect legal professionals’ practical challenges rather than relying on one-dimensional metrics.
Key points include:
– No large language model (LLM) guarantees 100% accuracy, necessitating a two-step verification process for legal research.
– Evaluating the entire research process rather than focusing solely on initial output accuracy is essential.
– A comparison between two hypothetical research tools illustrates the value of quicker verification times.
– Proper benchmarking should reflect the intended use and outcomes that are valuable to customers, especially in legal contexts.
As legal professionals emphasise, successful AI integration in law firms aims to enhance the research process without completely replacing human analysis.
Overall, GenAI should be assessed as part of a broader workflow that ensures thoroughness and accuracy in legal research.
Source: Above the Law
Leave a Reply