The rush to implement GenAI has left some law firms scrambling to integrate its use with pre-existing technology; but as a panel at ILTA Evolve revealed, a plan for evaluation, adoption, and iteration can go a long way
MYRTLE BEACH, S.C. — In the rush to adopt generative AI (GenAI) technologies, there is one simple fact that some eager attorneys and law firm leaders may forget: GenAI still won’t be the only technology in use. Indeed, any GenAI system not only needs to stand on its own, but it also needs to be integrated into pre-existing data systems, technology processes, and attorneys’ daily workflows.
Sound difficult? The reality of true integration may be even tougher. According to a panel, AI Made Practical: Enhancing Existing Tools and Driving Firm Adoption at the recent 2025 ILTA EVOLVE conference, any legal technologist entering the GenAI world with expectations of a perfect rollout will soon have those notions roundly distinguished.
However, that doesn’t mean that driving adoption of GenAI tools is impossible. It simply takes planning, from initial evaluation through implementation and feedback.
First, the evaluation
When the firm Frost Brown Todd began its GenAI journey, firm leaders started with urgency: They needed GenAI now, said panelist Masako Hashimoto, the firm’s Senior Legal Project Manager. “Our thing was that GenAI was going to change the practice of law, and if we don’t figure this out, others will,” she said.
This meant evaluating not only AI tools, but their applications to firm attorneys, all at the same time. Hashimoto noted that firm leadership understood that the GenAI implementation didn’t need to immediately be perfect, and it would be supplementing rather than replacing tasks. They simply required speed.
Many law firms have found themselves in the same conundrum, given the rapid pace of GenAI development since ChatGPT’s initial release less than 30 months ago. However, this need for rapid implementation means that many legal technologists have had to learn not only a new technology, but how lawyers would be implementing it in a practical way.
GenAI was going to change the practice of law, and if we don’t figure this out, others will.
Panelist Victor Chavez, Knowledge & Innovation Attorney at Sheppard, Mullin, Richter & Hampton, said he has seen some firms fall into the trap of evaluating technology before knowing how it’s going to be used — a totally backwards mindset. “You can have the coolest, shiniest AI car that can drive you anywhere you want to go,” Chavez said. “But if you don’t have a map — the problem you’re trying to solve — you’re not going to get anywhere effectively.”
Yet, even if legal technologists don’t immediately know the use case for an AI tool, they should be working on ways to solicit that feedback quickly. Another panelist, Christopher Fryer, Chief Information Officer at Hanson Bridgett, explained that when his firm was testing Microsoft Copilot, he saw quick interest among the human resources team — both because of the use case potential, but perhaps more importantly, because of weekly meetings the team would have to share worthwhile prompts and tips.
It’s an idea he’s taking forward to fully roll-out Copilot and other tools, with a monthly session in which attorneys can highlight to one another what they’ve learned. “One of the most important things you can do is gather those use cases and showcase them,” Fryer said.
Encouraging adoption
Even after a GenAI tool is selected, however, the work is far from done. Frost Brown’s Hashimoto said she approaches GenAI adoption as “not a technology project, it’s a change management project.” In one recent case, her firm’s attorneys were very excited about a new tool, and the firm was excited to provide training and licenses — until two months later, when Hashimoto noticed usage had dropped precipitously.
It was a case of mismatched expectations. “They thought that GenAI technology was something that would work like a washing machine. They would put the dirty clothes in, they would press the button, and they would get what they get,” Hashimoto said.
Undeterred, she and the Frost Brown Todd team retooled. They relaunched the tool internally with a team of 24 specifically chosen lawyers across various practice areas. The attorneys dedicated between two and four hours a week to dive into the technology; and crucially, the firm would give them up to 30 billable hours credit for doing so. Hashimoto’s team was there for white glove help, but not to tell people how to do things because they wanted attorneys to adopt the tool in ways that would fit with their pre-existing work.
“That vulnerability approach really worked,” she explained, adding that the attorneys would “learn from each other much better than the formal training sessions.”
This can make encouraging adoption even more difficult for legal technologists, noted Chavez of Sheppard Mullin. “The hardest part of this current period right now is that the technology is the easy side,” he said. “The harder part is making your knowledge and your mastery an approachable asset for the attorneys.”
They thought that GenAI technology was something that would work like a washing machine. They would put the dirty clothes in, they would press the button, and they would get what they get.
For his part, Chavez said that he’s spent much of the past year simply trying to be available within the firm, even to the point of being “annoying.” Even if attorneys learn best from one another, the technology function can still be a helpful ear in a time of need. Plus, simply by being available, firm technologists can more easily capture precious data.
“We created those spaces for the attorneys to come by and say, ‘I was thinking of doing 1, 2, and 3, how can AI help me?’” Chavez said. “And that’s where you collect your use cases.”
Supporting ongoing iteration
Indeed, law firm technologists may be operating with a dearth of data in the early days of the GenAI lifecycle. Hanson Bridgett’s Fryer acknowledged that not many have figured out the magic formula to measure GenAI’s ROI. For now, he’s focused on a path to building ROI in the future: Start small, build those wins, and have a story that people can rally around.
“I would not lead with this as an efficiency tool, because every time I do with a partner, they walk away,” Fryer said. Instead, he’s approaching GenAI an idea-generation tool. “In order to build ROI, you have to have a story, and you have to have something that people will listen to.”
In these early GenAI days, the panel suggested that it’s important for firm leaders to know where to focus their efforts. Hashimoto said she sorts attorneys into four general buckets: i) those who are skeptical that AI can perform any type of legal work; ii) those who don’t care and will continue to use their old processes; iii) those who are far ahead and know everything about the technology already; and iv) those who may be skeptical but are also interested in seeing where AI can help.
It’s with that last category where Hashimoto has found the most success. To continue to nurture adoption, she’ll make sure to schedule follow-up meetings with these attorneys to both see whether they’re using a tool effectively but also to ensure they know her team is available to help.
After all, her project management role — and the role of most legal technologists — is to respond directly to what their firm’s stakeholders need. And that maxim has not changed with GenAI, she said. “Every time you talk to an attorney interested in GenAI, really understand where they’re coming from and then speak from their standpoint of where GenAI can help them.”
You can find more of our past coverage from ILTA events here
