The legal profession’s commitment to eliminating risk may be a liability when it comes to successfully harnessing artificial intelligence for legal work.
A report from The Positive Group, developed with researchers from Harvard Business School, finds that even as technical barriers to AI use are falling, some law firms are resisting AI adoption for professional-cultural reasons.
The researchers called this the “zero-risk conundrum”: a profession devoted to maximum certainty is confronting tools built on probability.
A Shift in Legal Value
According to the report, more than 60% of lawyers are using AI for tasks like drafting, research and due diligence, compressing work that once defined early legal training.
“AI is fundamentally reshaping where expertise creates value in legal practice,” said Will Marien, CEO of The Positive Group. “As AI automates structured tasks, the premium moves to judgment — interpreting risk, navigating ambiguity and advising clients in complex situations.”
For most firms, the challenge is no longer whether to adopt AI but how to integrate it without eroding the quality of legal reasoning.
The ‘Liability Trap’
Firm leaders interviewed for the report repeatedly raised a growing tension: Clients are both pushing for AI-driven efficiency and demanding full accountability.
The result is what the report describes as a “liability trap,” or using tools that can generate probabilistic outputs in a professional world that expects accuracy.
“In a sector built on precision and accountability, AI introduces a new kind of uncertainty,” Marien said. “The primary concern is the professional and reputational consequences of getting it wrong.”
“Zero-risk isn’t realistic,” notes the report. “The challenge is to balance professional obligations with the need to move forward.”
Culture, Not Technology, as a Barrier
While much of the public conversation around AI has focused on hallucinations and data security, the report suggests the primary issue is cultural: Lawyers, trained to avoid error, may be reluctant to experiment with tools that introduce uncertainty.
AI adoption hinges on leadership behavior. “When partners and managing partners are visibly engaging with AI — testing it, sharing what works, acknowledging uncertainty — it creates permission for others to do the same,” Marien said.
Rethinking the Associate Pipeline
The AI effect is upending the profession’s traditional apprenticeship model. Tasks that once trained junior lawyers are increasingly automated. In their place, young associates are being pushed into higher-level work, including client advisory and analysis of AI-generated outputs.
Marien said, “The risk is that there is now less requirement to interrogate sources,” increasing the chance that AI-generated outputs are accepted without sufficient scrutiny.
In response, firms are placing greater emphasis earlier in lawyers’ careers on areas less susceptible to automation, such as critical thinking, communication and leadership skills.
From Risk Avoidance to Risk Calibration
Rather than drawing rigid lines around where AI can and cannot be used, firms can focus on building shared understanding across their team, exploring where AI adds value, where oversight is essential and how professional judgment applies in specific contexts, according to the report.
“The most effective approach is disciplined experimentation tied to real legal work,” Marien said. AI pilots with clear success criteria and decision points for scaling or abandoning projects perform best.
A Question for Leadership
The report ultimately frames AI adoption as a leadership challenge.
“Clients still value judgment, trust and accountability,” Marien said. “As AI becomes more capable, those qualities are becoming more, not less, important.”