AI Spending Surge Prompts Companies to Wrestle With Its Worth

In-house legal departments have bought artificial intelligence tools to churn through contracts, speed up workflows, and analyze documents. Now comes the hard part: Proving the software is worth the cost.

Some legal departments say AI software is improving how their lawyers work and boosting quality, but those enhancements can be difficult to quantify in a way that appeases the finance department. At the same time, some AI legal tech advocates worry that trying to demonstrate the tech’s return on investment will squander innovation that shouldn’t be constrained by concerns about the bottom line.

“GCs are caught between CEOs and boards wanting to know they’re using AI to find efficiencies, and concern and skepticism by the same people that they’re not making reckless acquisitions of technology,” said Eric Dodson Greenberg, executive vice president, general counsel, and corporate secretary of Cox Media Group.

The demand for ROI analytics is a sign of the maturation and proliferation of AI tools available to in-house lawyers. How much software legal teams are able to buy will depend largely on how effectively they can show benefits.

Legal teams are grasping for data points wherever they can find them, quantifying time savings, increases in output, and decreases in headcount. They’re leaning in to their clearest anecdotes about how AI is helping their teams.

Om Jahagirdar, deputy general counsel at Amtrak, said that legal departments are telling C-suites, “‘I was able to do X matters more. I was able to do this number of projects.’ And sometimes it might be hard to show that efficiency.”

Quantifying Impact

The push for ROI on AI investment comes from a concern that the technology hasn’t yet paid off. Even though up to $40 billion has been spent on enterprise AI, 95% of organizations aren’t getting any return on their AI investment, according to an MIT report published in July.

Rudy DeFelice, the global head of Harbor Labs, which consults on tech use, said trying to nail down ROI on AI would be like trying to measure the payoff of a desk chair or laptop.

“There’s certain things we just know we need,” DeFelice said. “I don’t know that anybody’s done an ROI analysis on any of that stuff.”

While return on investment is language that finance departments speak, legal teams are doing their best to boost their fluency.

Amtrak, for example, has asked attorneys to keep track of how long it takes to complete certain tasks using specific tools so they can show how AI has made them more efficient, Jahagirdar said. That’s a valuable data point to make the case for AI investment, he said.

“Otherwise, you’re just kind of using buzzwords, and they don’t know how to value that,” he said at the Association of Corporate Counsel annual conference in October.

Greenberg, at Cox, said his legal department reduced its headcount and offset it with a contracted attorney and an AI tool. The same amount of work got done, but at a clearly lower cost, he said.

Other AI benefits are harder to show. In a recent mediation, Cox relied on an AI tool to help craft its strategy, but it’s difficult to say how much of a direct impact the tech had on the positive outcome, Greenberg said.

“I have to be able to explain rather than show to the CEO that I used this tool and this is the benefit I got,” he said.

‘Theoretical Loss’

Some of the challenges to showing the return on AI are unique to legal departments.

“It’s hard for us to prove ROI anyway because we’re helping avoid risk, which is theoretical loss,” said Gabe Saunders, former head of legal operations at the professional coaching company Exos.

Legal departments might not generate revenue, but they’re not entirely divorced from it either, Saunders said. Exos’ legal team leaned heavily into contract playbooks, which sped up closing deals. Some contracts that once took six weeks to turn around could be completed in a few days, he said. And that resulted in more work done even with a slightly lower headcount, he added.

The contract management software that Exos used has a ticketing system that provides valuable data for demonstrating productivity, Saunders said. The more data the department has the easier it is to prove it’s spending wisely, said Saunders, now legal operations strategist at LegalSifter, which makes contract management and review software.

Legal operations leaders also play a key role, said Monica Zent, founder of an alternative legal services provider who consults with in-house departments. Those professionals have for years been tasked with vetting software and making the case for tech adoption, she said. Now, they’re applying the same skills to AI.

“Not only the C-suite, but even boards, corporate boards, are now asking for ROI. ‘Show me the ROI on our AI adoption here. Show me the ROI on what we’re what we’re buying,’” Zent said.

Return on Experience

Everything can’t be quantified, which is why some in operations point to measuring “return on experience.” That metric focuses on how AI improves an attorney’s job.

It might not show up in the bottom line, but general counsel still want their teams to be interested and engaged, said Stephanie Corey, former HP Inc. head of legal ops. Getting rid of busywork to let lawyers focus on the more interesting parts of their job is a clear benefit of AI, she said.

“Nobody wants to be running through contracts and extracting key terms,” Corey said. “That’s not fun.”

AI is changing rapidly, which affects how companies work and what in-house teams can do. Constraining that impact by asking for numbers to back it up could prevent companies from achieving AI’s full benefits, Harbor’s DeFelice said.

“Some guy in finance requires some kind of ROI, but it’s not a courageous way to think about innovation,” he said.