I spent 17 hours at a legal conference. These were the 2 questions haunting lawyers about AI.

• At Legalweek, AI was a hot topic, but many lawyers remained skeptical about its implementation.
• Concerns over job security and inadequate training hindered their adoption of legal-tech tools.
• Billions of dollars are riding on lawyers moving faster with tools from startups Harvey and Legora.

At Legalweek this week, lawyers got a mix of shmooze, swag, and one topic on repeat: artificial intelligence. It headlined nearly every panel and powered every booth pitch. Still, the question I heard most — in hallways, at happy hours, even from the people selling the tools — was disarmingly basic: How do we get lawyers to use it at all?

What's happening in Big Law looks increasingly like what's already unfolding in tech. Companies like Atlassian and Block have already tied recent layoffs to AI-driven efficiency gains, signaling that the technology's promise of higher productivity can also mean fewer jobs.

Billions of dollars are now riding on the assumption that AI will do the same for professional services. But inside law firms, adoption is a mixed bag, even as clients demand faster, cheaper work and investors expect returns on the billions of capital flooding into legal tech.

Legalweek, held at New York City's Javits Center, has become an annual progress report on whether generative artificial intelligence is actually changing an industry that's cautious by design. It was my second year attending, and the demos were noticeably slicker. AI "agents" were everywhere across the expo hall, sold as digital coworkers that can draft, review, and run multi-step workflows that used to require a junior associate and a lot of coffee.

But the conference also produced a quieter signal that the adoption curve isn't matching the hype. When Microsoft's Steven Abrahams, who leads product work on Copilot integrations, asked a room of attendees to raise their hands if they use software to automate contract review — one of the clearest use cases for large language models — only a few hands went up.

Onstage, the warnings were blunt: Clients will take their business elsewhere if firms don't change their habits. "Revenue is at risk," Emma Dowden, Burges Salmon's chief operating officer, said during a panel hosted by Harvey, an $8 billion startup selling software to law firms and corporate legal teams.

In that same session, a moderator asked Derek Morales, an in-house lawyer at Macquarie Capital, whether "AI maturity" will factor into how companies choose outside counsel a year from now. "I'm judging today," he said, adding that it's "cringeworthy" to hear firms say they hired a new chief innovation officer while balking at buying licenses for legal artificial intelligence platforms.

All week long, lawyers tried to explain the tepid response. Dowden pointed to fear. Lawyers, she said onstage, worry about what automation means for their own jobs, how it could eat into work billed by the hour, and whether they understand the technology well enough to defend it to a skeptical client. The anxiety can curdle into resistance. Partners may want the tech's upside, she suggested, as long as another practice area in the firm tests it first.

People often assume younger lawyers will be the easiest converts. Not always, said Sarah Eagen, who leads learning and development at the megafirm Cleary Gottlieb, during a panel. Cleary Gottlieb has rolled out Legora, a Harvey competitor, firmwide. Still, many associates see automation as a threat, Eagen said, just as they've spent years and money buying into a career built on entry-level work.

Lawyers aren't alone in this anxiety, as Business Insider's Tim Paradis has written. In industries where AI adoption is further along, job cuts are routinely being framed as the natural next step — the cost of doing business in an era where work moves faster, and companies insist they simply need fewer people to do it.

Lawyers are trained to make decisions only after they've gathered the facts. Speakers suggested that when firms actually invest in training, lawyers are more likely to use the tools — because they understand the guardrails.

Ian Nelson, who runs Hotshot, a company that helps law firms build training programs, said during a panel that too few firms offer any AI training at all. "There seems to be this mentality" that training can wait until a firm has licensed a tool, he said. That's short-sighted, Nelson argued, because some lawyers will use chatbot tools anyway. And when training does happen, he said, it's often too narrow. Think tool-specific demos without the context about risks and the firm's own policies.

Legalweek's loudest question was still the one making vendors squeamish: Why won't lawyers use the tools? But as the week went on, a second question started to surface.

If AI really can produce better, more cost-effective service, at what point does resistance start to look like malpractice?

Corporate lawyer Michael Pierson raised the issue during Harvey's panel. Just over two years ago, Pierson and his fellow former FisherBroyles partner Joel Ferdinand broke from the firm to launch Pierson Ferdinand, a smaller, distributed outfit that leans heavily on tools like Harvey and runs without any associates.

"If we're not using AI in the daily delivery of our legal services to our clients, is it malpractice per se? I don't know," he said. "But we're in the business of client service, and anything that gets us to excellent work product, we should be exploring."

Have a tip? Contact this reporter via email at mrussell@businessinsider.com or Signal at @MeliaRussell.01. Use a personal email address and a non-work device; here's our guide to sharing information securely.