Your Firm May Already Have an AI Policy. It Just Has Not Been Written Down.
Many law firms still speak about AI adoption as if it were a future decision. That framing is becoming less accurate by the month. In many firms, AI has already entered the practice, not because leadership held a retreat and approved a strategy, but because lawyers and staff are making practical choices inside ordinary work. They are using AI to draft correspondence, summarize documents, brainstorm, edit language, create templates, and accelerate research. Vendor platforms are also adding AI features into software the firm already uses. The result is that a firm may already have an AI adoption pattern, even if leadership has never deliberately chosen one.
The "2026 8am Legal Industry Report" brings that pattern into view. It found that 69% of legal professionals surveyed personally use general-purpose AI tools for work-related purposes. At the same time, 54% said their firm has provided no responsible-use training and has no current plan to do so, while 43% said their firm has no formal AI policy and no plan to develop one. Those numbers should not be treated as a perfect portrait of every mid-size firm. The respondent pool was heavily weighted toward solos and small firms. Even so, the leadership pattern is difficult to ignore. Individual use is moving faster than firm-level visibility, training, supervision, and governance.
That does not mean firms are acting recklessly. Many are cautious for good reasons. Data security, privilege, confidentiality, supervision, and trust in the output are all serious concerns. A prudent managing partner should hesitate before allowing AI-assisted work to spread casually through client matters. But that is precisely why visibility matters.
A firm can be cautious at the policy level while becoming experimental at the workflow level.
Leadership may believe the firm has not yet adopted AI, while AI is already shaping how work gets drafted, summarized, reviewed, or prepared.
That quiet mismatch is the real management issue. It is easy to imagine AI as a matter of formal tool deployment, software procurement, or future policy drafting. In practice, AI often arrives through individual behavior before it arrives through institutional decision. One lawyer uses a general-purpose assistant to revise language in an email. A staff member summarizes a long document with an AI feature built into an existing platform. A paralegal experiments with drafting a template or cleaning up a chronology. None of these moments looks like a grand strategic shift. Taken together, they can become the firm’s actual posture.
Professional responsibility does not wait for leadership to finish discussing the issue. The work still carries the firm’s name. The obligations of confidentiality, supervision, competence, and judgment still apply. That is why the first question for a managing partner is not whether the firm should join the AI race. The first question is whether leadership can see clearly enough to describe the firm’s current posture. What is already happening inside the practice? Which uses are low-risk and appropriate? Which require closer review? Which should remain off-limits until the firm has made a more deliberate decision?
Responsible adoption begins there. It does not begin with a software demo, a policy template, or a vendor presentation. It begins with visibility. A managing partner needs a clear picture of where AI is already touching the firm’s work and what assumptions people are making about acceptable use. Only then can leadership decide what belongs inside the firm, what requires supervision, and what should be restricted or prohibited for now.
That kind of decision-making does not require panic, and it does not require grandiosity. A firm does not need a sweeping transformation initiative to get started. It needs a practical leadership posture. Someone has to decide what kinds of AI use are acceptable, what information may be entered into which systems, what output must be verified before it is relied upon, who is responsible for review, and how the firm will communicate those boundaries to lawyers and staff. The work is not glamorous, but it is foundational. Without it, informal practice will continue filling the gap.
The deeper risk is not only that someone may use ChatGPT badly. The deeper risk is that firm leadership may not know how many small decisions are already being made below the level of formal policy. Those decisions accumulate. Over time, they shape habits, expectations, workflow norms, and eventually client-facing work. By the time leadership looks up and says, “We should probably decide what our AI policy is,” the firm may already have one in operation. It just may not have been written down, supervised, or intentionally chosen.
That is why I think the AI adoption question is already inside many law firms. The issue is no longer abstract. It is not reserved for innovation committees, legal technology conferences, or vendor webinars. It is a present leadership question about visibility, judgment, and professional guardrails. Firms that address it early can make deliberate decisions before default behavior hardens into firm practice. Firms that wait may discover that their real posture was set one quiet workflow at a time.
For managing partners who want a clearer view of their firm’s current AI posture, I prepared a short self-assessment for law firm leadership. It is designed to help surface where AI may already be present, where leadership has made deliberate decisions, and where the gaps are still open. No sign-up is required.
Rex C. Anderson
The retired attorney law firms call about AI Adoption
Forwarded to you?
Subscribe for direct delivery of future articles.