Litigators at Massachusetts firms are weaving artificial intelligence into daily tasks-document review, legal research, drafting pleadings, and managing discovery-but they treat the technology as a starting point, not a finished product. "It's a great check on your work," said Sean Carnathan, a trial lawyer at O'Connor, Carnathan & Mack, "but it's not a replacement for having a human brain that knows how the law works."
A starting point, not a finished product
Carnathan feeds briefs from opposing counsel into an AI tool alongside his own draft and asks it to sharpen his arguments. He also runs deposition transcripts through summary software like Clio. The output is useful but can mislead. AI doesn't always understand how a case applies, he said; it might pull a quote out of context and present it as the controlling rule. "It will divine which side of the 'v' you're on, and tell you what it thinks you want to hear," Carnathan said.
Nicole Benjamin, a trial and appellate litigator at Adler, Pollock & Sheehan, noted that her firm uses one of several legal-specific AI products-Lexis Protégé, CoCounsel, and Harvey-and is piloting five more. Benjamin refuses to let the technology reach remote depositions; she wants to gauge a witness's demeanor in person. She also spotted fabricated evidence that an AI generated because she knew how the tool behaves. A lawyer without that fluency, she said, could have accepted it as real.
Billing shifts and ethical obligations
The technology is already changing client conversations. One client repeatedly sent back AI-generated versions of work, questioning why an associate billed for what a machine did in minutes. Carnathan moved that client to a flat fee. He isn't ready to bury the billable hour-its death has been predicted for decades-but added, "we might finally have a technology that's going to corner lawyers into changing." Pricing for the tools is still unsettled, said Howard Goldberg, a trial lawyer at MG&M and chair of the firm's e-discovery committee. Costs swing from a modest expense for a general assistant to $1,200 a seat for enterprise systems.
Goldberg stressed that the duty of competence now demands understanding AI. "AI is exciting and holds enormous promise for the profession," he said. "However, attorneys who use AI are required to understand the technology, educate themselves about it, and ensure that its output is accurate." He compared the shift to the adoption of e-discovery a generation ago-it isn't optional. "Your duties to your client and the court have not changed. If you use AI, you must educate yourself on how the software works and where it may fall short." Lawyers can build that knowledge through structured training in AI for legal professionals, which covers the limits and reliability of the tools now appearing in litigation workflows.
Hiring priorities for new litigators
With AI capable of producing clean prose, the skills firms value in new associates are shifting. "It's not replacing anybody, but it's very much changing the skill sets that we rely upon," Benjamin said. She now looks for interpersonal abilities-managing client expectations, knowing when to push a settlement, delivering bad news-and analytical strengths like spotting the weakness in an opponent's argument. These once developed over time. "Now it's really critical that you have that coming in," Benjamin said. Carnathan described the change bluntly: lawyers will either ride the wave or get "crushed on the beach."
Why this matters for legal professionals
AI won't eliminate the billable hour overnight, but it is forcing firms to defend the value of human judgment and to rewrite billing arrangements. Litigators who treat the tools as a check, not a crutch, and who invest the time to understand how the software can fail are the ones most likely to protect their clients-and their own practices. The duty of competence now includes a working knowledge of AI output, and the hiring market is rewarding those who bring that awareness from day one.
Your membership also unlocks:
