3 Generative AI Impacts E-Discovery Professionals Are Watching Closely

As e-discovery professionals seek to understand the specific impact of generative AI, some considerations are more pressing than others.

As the legal industry works to understand the impact of generative artificial intelligence on their particular sectors, it is finding a handful of aspects may likely change, while others could stay the same.

This is especially true for e-discovery professionals, who are likely to deal with the technology in various ways—as evidence, as a tool to power discovery, and as considerations in disputes.

Below are three potential generative AI impacts that e-discovery professionals are keeping an eye on.

Will Generative AI Reduce or Increase E-Discovery Costs?

Few aspects of any new technology are as important as its price tag. While many in the e-discovery industry believe generative AI is capable of automating time-consuming tasks and adding value to the sector, experts are split on whether this is enough to stomach the technology’s upfront cost.

For instance, Sidley Austin’s counsel of data analytics and discovery, Matt Jackson, believes that generative AI’s upfront cost is far too high to justify its use for many of their clients.

“We saw this with TAR early on, [that] any new technology is going to start at a higher price point and then over time technology gets cheaper and better,” Jackson said. “But I represent my clients. My clients are cost-conscious. I can’t go to them and say, ‘I want to use this really nifty, cool tool, but it’s going to cost you more. That’s never going to fly.’”

To be sure, not everyone feels like the initial sticker shock is necessarily the death knell for the technology. Michael Cichy, the litigation support general manager at Foley & Lardner, said that the long-term use of generative AI in e-discovery will likely lower costs. For example, since many of the tools powered by the technology come built with the capability to offer “reasoning” behind why they are generating or flagging certain documents, it can help save time, Cichy said.

“So even though the initial sticker shock might be there, down the road, it might still wash out and might wind up getting some downstream savings,” he noted.

Will ESI Protocol Change to Accommodate Generative AI?

Lawyers can do their best to understand generative AI and apply the necessary safeguards to their client’s data when using the technology. However, they can’t necessarily be sure of the same from opposing counsel.

While some believe that the existing ESI protocols may be enough to accommodate the unique risks of generative AI, others aren’t so sure. Sandra Metallo-Barragan, an e-discovery counsel at Proskauer Rose, argued that attorneys that try to ban the opposing party’s use of the technology are likely going too far, because existing ESI protocols are flexible enough to adapt to the growing technology within legal.

Still, others believe ESI protocols will need to evolve. Not unlike stipulations attorneys put forth in the early days of technology-assisted review (TAR), around search terms or certain restrictions, Metallo-Barragan said questions around which large language model specifically is being used and what the data storage properties of the tool look like, among others, should be in ESI protocols.

But, like any technology, after a certain point, opposing counsel simply have to be trusted to be tech-competent. At the end of the day, micromanaging generative AI use too much through ESI negotiations is like going to cause longer e-discovery disputes, she noted.

Will Generative AI Have an Impact on the FRCP?

An inevitable consideration that comes with any new technology is how it fits into the tried-and-true framework of the law.

Increasingly, attorneys, courts and judges are considering whether the existing Federal Rules of Civil Procedure (FRCP) will have to be amended for the age of generative AI.

William Eskridge Jr., Alexander M. Bickel professor of public law at Yale Law School, said it’s more likely than not that the technology will have an effect on certain e-discovery rules.

“It’s clearly going to transform the discovery rules like Rule 26(b)’s proportionality requirements,” Eskridge said at the D.C. Bar Summit. The debate around this rule consists of how parties determine what discovery is “proportional” to the particular needs of a case.

As the technology increases the ability of e-discovery teams to go through documents as a faster rate than before, the debate around proportionality is primed to evolve.

“Proportionality still has a place in the rules,” said Nicole Gill, an e-discovery attorney at Cozen O’Connor. ”Now, will that change as, say, generative AI makes it easier to go through thousands, and tens of thousands, hundreds of thousands, and sometimes even millions of documents to look for very specific nuances or ideas or relevant documents?”