LawGeex, an Israel-based startup focused on automating contract reviews, released a study showing its deep learning software outperforms lawyers at identifying legal risks in nondisclosure agreement contracts.
“Don’t expect machines to kill lawyers’ careers though,” said Shmuli Goldberg, vice president of marketing at LawGeex. He likened the use of LawGeex AI to people using spellcheckers in their emails or pilots using autopilot on planes, both enhancing their work.
Using Tesla P100 GPUs on the Amazon cloud and the cuDNN-accelerated TensorFlow deep learning framework, the team trained their neural networks on tens of thousands of NDAs. The trained software reached 94 percent accuracy in surfacing risks in previously unseen NDA contracts compared to the lawyers used in the study who had an average of 85 percent.
The AI was also several hundred times faster at the job – they did the NDA reviews in only 26 seconds, versus a span of about 1 to 2.5 hours for the lawyers.
“Lawyers are realizing this actually can make their job better,” Goldberg said. “This work is the everyday gruntwork, and most lawyers can’t wait for an AI to take it off their plate.”
Read more >
Related resources
- GTC session: Model Development for Bias, Factuality, and Attribution
- GTC session: Intellectual Property Challenges in the Age of Generative AI
- GTC session: Best Practices for Building LLM RAG Using NVIDIA AI
- NGC Containers: NVIDIA Retrieval QA Mistral 4B Reranking v3
- NGC Containers: NVIDIA Retrieval QA Mistral 7B Embedding v2
- NGC Containers: ASR Parakeet CTC Riva 1.1b