In one of the most recent ‘man v machine’ battles, lawyers were pitted against an AI algorithm. It turned out that AI achieved an average 94% accuracy rate, while the humans came in at 85%.
The US-trained lawyers who took part aren’t novices: they have decades of experience in corporate law and contract review. The task at hand? To identify legal issues in five standard NDAs (non-disclosure agreements).
The company behind the algorithm is LawGeex and it’s been developing the AI system for three years, training it on tens of thousands of contracts.
Shmuli Goldberg, LawGeex’s marketing VP told MarketBrains that the AI is tweaked to review NDAs specific to financial transactions as well. Clients include major French financial services firm Natixis and LA-based investment firm Skyview Capital.
Financial-specific NDAs seem to be a real sweet spot for an AI approach because they are high volume standardized contracts, but specialized to a sector.
“They (financial services firms) use us as a triage tool for looking at contracts in the first place,” Goldberg explained.
“By using an AI tool to get past a level-of-expertise barrier, they are starting to use people with a wider range of skills that may not have had experience in this specific type of contract.”
He noted that the legal sector is catching up to finance in terms of automating as much as possible.
“The reason we are seeing success with financial customers is because they have this innate belief that if I can automate it, I should automate it, and that’s not language traditionally lawyers come along with,” he said.