-
AI has been used in legal circles for some time, but GenAI is enabling a new range of use cases
-
GenAI has the potential to save law firms and enterprises thousands of hours and dollars by tackling menial tasks like document review, contract drafting and more
-
Companies tackling GenAI for legal services say adoption is accelerating
When I was a kid, my attorney father had a shirt with a modified version of that famous line from Shakespeare: “The first thing we do, let’s kill all the lawyers.” Instead of “kill” the shirt said “kiss.” A clever change. The line is commonly seen as belittling lawyers, but some argue it actually shows their importance. The same could almost be said of the artificial intelligence (AI) that is being used to tackle certain enterprise legal tasks.
Rather than trying to replace lawyers, new generative AI tools from ContractPodAI and others aim to let lawyers delegate repetitive menial tasks and focus on the legal work AI can’t do.
According to a recent NYU Law Forum panel, lawyers have actually been using AI for quite some time to help with the discovery process – that is, figuring out which case documents you have to turn over to the other side and reviewing the ones the other side gives you.
Betny Townsend, product marketing director for DISCO, a cloud-native provider of AI for legal services, said during the panel that legal AI to date has basically functioned like a sophisticated Netflix recommendation algorithm. Basically, it has been used to recommend which documents in the discovery stack they might want to review in detail.
While that can make attorneys “significantly more efficient, it still requires a lot of human input,” she said.
Generative AI, however, takes things to a whole new level. Not only can large language models sort through millions of documents to find ones that meet a certain set of criteria, but they can also respond to queries as well as write and summarize deposition transcripts, Townsend said.
That’s great for lawyers, but Anna Gressel, counsel at law firm Paul, Weiss, Rifkind, Wharton & Garrison, noted enterprises can also use those tools to discover information buried in decades old corporate documents for contractual and compliance use cases. For example, AI can help figure out what kind of language has historically been used for a certain kind of contract or surface relevant policies. Those are precisely the kinds of legal-adjacent tasks AI is taking over.
“These are companies that are global. They may have hundreds of policies that could be applicable to a certain situation,” Gressel said.
GenAI can help them find the right policy quickly. She added GenAI is also being used to assist with internal investigations into employee grievances.
Kill all the lawyers?
According to IDC data, spending on generative AI solutions is expected to skyrocket from $16 billion in 2023 to $143 billion in 2027.
Asked in an interview with Fierce whether GenAI is going to take over the world and replace lawyers altogether, ContractPodAI CPO Atena Reyhani laughed and said no. However, as noted above she said it will – and already is – enormously streamlining legal workflows.
ContractPodAI (which competes with the likes of IronClad, Icertis and others) offers a sort of GenAI legal assistant called Leah. There are 10 different out-of-the-box Leah modules that have been customized for specific legal tasks – everything from discovery and data extraction to document drafting, deal negotiations and providing HelpDesk-style guidance.
Leah launched last year and Reyhani said enterprises spent much of 2023 experimenting with the tech. This year, though, has seen a significant uptick in actual adoption, particularly in Q1, she said. Reyhani said customers increasingly include law firms and clients referred through partners like Google Cloud in addition to converts from ContractPodAI’s legacy software-as-a-service customer base.
Notably, professional services and accounting firm PwC recently announced plans to offer a consulting service based on Leah.
At the beginning, Reyhani said questions and concerns were focused on security and privacy, as well as addressing issues like hallucination (which she said ContractPodAI does on the architecture layer with built-in engineering tools like query expansion and retrieval augmented generation mechanics).
Now, though, the conversation has shifted to relevance and time to value.
“Basically, for customers – especially on enterprise-grade products – the question is ‘can GenAI give me recommendations and analysis that is relevant to me as an organization with my rules, with my preference and negotiation and strategies?’” she said. “I think that is a new area of conversation.”