Turning Legal AI Strategy into Action: Challenges and Opportunities

Insights from leaders at Honigman’s 2025 Innovation Symposium

Publication

Legal AI is at a pivotal moment. Nearly a third of legal teams are using AI and more than half plan to adopt AI tools in the next two years, according to a 2025 survey from the Corporate Legal Operations Consortium. Yet integrating an AI tool is often a complex and challenging process, leaving teams to grapple with how to turn ambitious AI strategies into real, measurable results.

To understand how both new and mature adopters are leveraging legal AI in their workflow, Honigman brought together legal and corporate leaders at the 2025 Innovation Symposium this September. Here’s how their teams are responding to implementation challenges, as well as the promising opportunities on the horizon.

Challenges to Legal AI Deployment

Emerging technologies like AI come with a host of issues that can present roadblocks during implementation, including:

The People Problem

Cultural inertia can be one of the most significant hurdles for any new technology, since people tend to prefer the tools they already understand and know how to use. This is especially true for busy and risk-averse lawyers.

“It’s hard to get people stuck in their ways to try something new,” said Emily Dillingham, corporate counsel for litigation and regulatory at Kellanova. “We are all still using the same platforms we’ve been using for far too long, even though there are much better tools on the market.”

It’s imperative, then, that leaders make a case to users about AI’s benefits. Lawyers who understand the future payoff in efficiency or risk management, for example, may be more willing to invest their time and energy in piloting AI now. If resistance to AI persists, consider working from the bottom up to select tools and develop processes that respond to their existing needs.

“Don’t push them a product, give them a process,” said Kevin Smith, Founder and CEO of IQI Balanced Intelligence, a construction and technology conglomerate. “Ask them what they would like to use. What would make their day better? Start there.”

The Process Problem

Understanding existing workflows—and how AI slots in—is a must for successful AI adoption. Otherwise, automation may even exacerbate current problems.

“The number one thing for us was not realizing just how much we would need to reimagine workstreams,” said Travis Rogers, director and assistant general counsel at Allstate.

Thus, rather than starting with areas that had not yet been reimagined, his team pivoted to an existing process involving state regulatory compliance. This workstream had already been mapped out for an earlier overhaul, so seeing where AI could add efficiencies was much easier.

Organizations could also consider giving lawyers the freedom to design their own preferred AI solutions. “We teach our teams how to use AI so they can solve their own problems, and we encourage that,” said Arjun Mehra, strategy lead at Google.

Rogers agreed: “What we've really leaned in on is finding the use cases from the users. We’re allowing them to use things like copilot for their daily work and asking ‘Is there some sort of efficiency you could find?’ instead of a top-down approach.”

The Ethics Problem

It’s no secret that AI has been used unethically in the legal industry, with “ghost citations” in AI-generated briefs or even legal decisions leading to sanctions. AI tools, particularly general-purpose generative AI platforms like ChatGPT, can also present data security and privacy risks.

This means it’s critical for organizations to impose guidelines and encourage transparency. Rogers’s team, for instance, created guidelines that promoted safe AI use throughout the legal department.

In-house teams expect the same of their outside counsel partners. “I don't want my outside counsel to have to come to me and say I got sanctioned, and the penalty is that the judge said I had to tell you, and that also I have to pay a bunch of fines and report this to the bar,” said Dillingham. At the same time, she noted that in-house teams are often open to collaborative AI use cases that their outside counsel want to explore, provided the appropriate guidelines are in place.

Opportunities for AI Innovation

What makes AI implementation challenges worthwhile? The potential to revolutionize workstreams and unlock new potential. Some of the most promising applications for AI in legal are:

Automating Routine Work

A powerful advantage of legal AI is automating routine tasks that require low sophistication but high time investment to free up more bandwidth for complex work.

“I’m excited for the time when AI will allow me to concentrate more on enterprise-level thinking by taking some of the more rote aspects out of the job,” said John Knowles, associate general counsel at Old World Industries, automotive fluids and parts company. “We have a lot of really cool ideas that get bogged down in the day-to-day, and so AI’s ability to give me time to do that higher level thinking is exciting.”

The panelists explained how their companies are exploring AI automation, with use cases such as:

• Conducting preliminary reviews of contracts to look for inconsistencies or areas of risk.
• Analyzing new privacy laws across different jurisdictions to see how products will be impacted.
• Performing document reviews to tag materials and identify relevant files.

While this might take away some of the work done by early-career associates, the panelists noted that this isn’t necessarily a bad thing—so long as they receive other forms of hands-on training. In fact, automation can have an unexpected benefit for those just starting out: more job satisfaction.

“If your associates are happier, they're going to do better work for your clients, and that's better for everyone,” said Dillingham. “If they're not sitting in their office or at home all day doing painful deposition summaries, they're happier and they have more time to get real experience.”

Seeing New Patterns and Connections

More sophisticated and mature AI applications have the potential to yield new insights and communicate ideas in creative, engaging ways.

For example, Smith said his company is creating multi-dimensional data repositories (also known as knowledge cubes) that absorb everything from contracts and insurance policies to litigation documents and meeting transcripts. With all this information in one place, users from different sides of the business can use ChatGPT to search for information and identify issues.

“We’re going to see patterns we never saw before with AI,” he said. “It’s going to really change how humanity sees the world.”

AI can also help bring together information and context to communicate important information. “Powerful context is so hard to find at a company our scale,” said Mehra. “AI's ability to connect dots and bring together context when we have 1200 people in our legal department is so important.”

Charting a New Course for AI in Legal

AI’s impact on the legal industry is rapidly transitioning from theoretical to practical. Successful adoption depends not just on the right tool, but cultivating the culture, processes and guidelines needed to foster real innovation. For those who build a strong AI foundation, the opportunities could be game-changing.