I got a call last month from a CMO at an Am Law 100 firm. She had just walked out of a meeting with her managing partner, who told her that AI adoption was the firm's top strategic priority for 2026. Then he asked her one question: "What is your team doing with AI?"
She froze.
She called me from her office an hour later. Her team had no sanctioned AI tools, no internal training program, and nothing in the budget for either. She needed to get up to speed on AI herself, and she needed to hand her managing partner a coherent story about what her team was already doing. Both things needed to happen fast.
She asked if I would come in and run training for her team. I said yes. A training session was the one piece she could authorize on her own. The rest of what her firm needed would require a longer conversation, one she had not yet had permission to start.
That conversation has been replaying in my head ever since, because I have had some version of it with marketing leaders at nine different firms in the past six weeks. The details change. The dynamic stays identical.
Leadership declares AI a priority. A task force forms. Memos circulate. And then the marketing team receives nothing. No budget and no training on workflows. Just a directive and a generic chatbot.
I started calling this the AI Mandate Gap: the distance between what firm leadership says about AI and what they fund for the teams expected to use it.
The sequence is remarkably consistent across firm size, geography, and management structure.
It starts with a pronouncement. The chair or managing partner announces that the firm is committed to AI. Sometimes this happens at an all-hands. Sometimes it is a paragraph in a strategic plan. Either way, the message is clear: we are an AI-forward firm.
Next comes the task force. A committee of partners, IT leadership, and sometimes a few senior business professionals gets assembled. They review vendors. They discuss governance frameworks. Months pass.
Meanwhile, the practice groups get resourced. Lawyers receive access to Harvey, CoCounsel, or Legora. Custom integrations get built. Training sessions are organized. Partners get one-on-one onboarding. This makes sense because legal work is the firm's revenue engine, and the ROI on lawyer-side AI tools is straightforward to calculate.
The marketing and business development team gets Copilot. Maybe ChatGPT Enterprise. A blank prompt box with no knowledge of the firm's brand voice, client roster, competitive positioning, or editorial standards.
Back to the CMO who called me. Her firm had spent north of $400,000 on AI tools for lawyers in 2025. Her marketing team's AI budget was $0. Absent from the budget entirely, with no line item anywhere.
Three forces keep the mandate gap in place, and few firms have examined any of them.
The ROI visibility problem. When a lawyer uses AI to draft a brief 40% faster, the calculation is direct: billable hours saved, multiplied by rate. When a marketing team uses AI to produce a competitor intelligence report or a client alert, the value is tangible but harder to quantify in a single line item. Firms default to funding what they can measure, which means the legal side gets resourced and the marketing side gets told to figure it out.
The "tools are enough" assumption. Many firm leaders believe that providing access to a general-purpose AI tool is the same as enabling AI adoption. Giving a marketing team ChatGPT without workflows is like giving someone a commercial kitchen and expecting them to run a restaurant. The tool is a prerequisite, but it is nowhere near sufficient. Purpose-built workflows with firm context, brand voice parameters, and structured outputs are what turn a chat window into a productivity engine. Without them, you get twelve people on the same team independently building the same basic prompt for the same basic task, each in their own browser tab, with nothing learned at the organizational level.
Misallocated caution. Firm leadership is understandably careful about AI. They worry about confidentiality and reputational risk. But that caution gets applied asymmetrically. The legal side receives governed, vetted, purpose-built tools with guardrails. The marketing side gets consumer-grade products and is expected to build their own guardrails. The ungoverned approach is the one that creates more risk. When every person is improvising with a general chatbot, there is no consistency and no audit trail.
Firm leadership tends to treat the mandate gap as a minor inefficiency. The costs compound in ways that stay invisible until the damage is done.
Talent attrition. Marketing professionals at law firms are watching their peers at agencies, corporations, and competing firms adopt AI in meaningful ways. The best people in this field want to work with modern tools. When a senior director is told AI is a priority but given no resources to act on it, she starts looking. I have heard this from three different people in the last quarter alone: "I am going to leave if I have to spend two more years doing this manually when I know there is a better way."
Competitive positioning erosion. The firms that are operationalizing AI in marketing are producing higher-quality competitive intelligence and more responsive client communications. Their RFP responses are sharper. Their pitch materials are more current. Their client alerts go out the same day a regulatory change drops instead of three days later. The firms still running manual processes are falling behind in ways that are invisible until a client defects or a pitch gets lost.
Opportunity cost at scale. A marketing team of ten people spending five hours per week on tasks that AI could reduce to one hour is losing 40 hours of labor per week. Those are 40 hours that could go toward strategic work: client relationship building and proactive business development. Multiply that across a year and you are looking at roughly 2,000 hours of strategic capacity left on the table.
Organizational learning stagnation. When individuals improvise with AI independently, the firm learns nothing. Nobody knows which prompts work. Nobody knows which workflows produce reliable output. Nobody is building institutional knowledge. The firm stays perpetually at day one of AI adoption, even if individual team members have been experimenting for months.
I work with firms that have moved past the mandate stage, and the ones getting results share a common approach. It requires deliberate decisions, but the playbook is clear.
They start with two workflows. Firms that stall are the ones that try to "AI-enable everything" at once. In my experience, the better approach is to pick two specific, high-frequency tasks where AI can save measurable time. Staff meeting prep and competitor monitoring are strong starting points, as are client alert drafting and event follow-up sequences. They prove the value on those two workflows before expanding.
They invest in purpose-built tools. A blank chat window with no knowledge of your brand voice, competitive set, or client base will produce generic output every time. The firms getting results use tools that have firm context baked in. Systems that know the difference between your firm's approach to data privacy and your biggest competitor's. Tools that can draft in your house style, that reference your client roster when building targeting recommendations.
They fund it. One workflow subscription costs less than a single junior associate's weekly billing rate. A comprehensive AI toolkit for a ten-person marketing team runs less than a first-year associate's monthly compensation. When you frame the investment against what the firm already spends on legal-side AI tools, the disparity becomes difficult to justify.
They train on specific workflows. The AI training at law firms tends to be tool training: here is how to log in, here is how to write a prompt, here is the content policy. That is necessary but insufficient. The firms that build real capability train their teams on step-by-step processes for specific tasks, with templates, quality checks, and feedback loops built in.
If you are a marketing leader reading this and recognizing your own firm in the CMO's story, you already know the disparity exists. The work is in building the business case to fix it.
Start with a small pilot. Pick one workflow that your team performs weekly and that leadership cares about. Competitor intelligence is a strong candidate because partners notice when it is good and complain when it falls short. Run the pilot for 30 days with a purpose-built tool. Document the time saved. Document the quality improvement. Put specific numbers in front of leadership.
Frame it as risk reduction. The ungoverned approach, where twelve people freelance with ChatGPT, is the exposure. A workflow-based approach with firm context and quality controls is the mitigation.
Compare the investment to what the firm already spends. If the firm allocated $400,000 for lawyer-side AI in 2025, ask what the marketing team's equivalent looks like. The answer can be $15,000. It can be $30,000. But leaving it at zero is a deliberate choice to under-resource the team responsible for the firm's external voice.
And be direct about what the current state produces. An unfunded mandate produces frustration and a widening distance between what the firm says publicly about innovation and what it does internally.
I ran the training for the CMO's team. It went well. Her team was engaged, the questions were sharp, and people left with real skills they did not have that morning. But a half-day workshop does not replace what her firm had failed to build. She still walks into that meeting with her managing partner, and the managing partner still asks what her team is doing with AI. The answer gets a little harder to improvise every week.
Your marketing team wants to use AI. They are ready. They are waiting for you to make it possible.
The mandate gap is a resource allocation problem. And unlike most resource allocation problems in law firms, this one has a straightforward fix: pick the workflows, fund the tools, train the team.
Every quarter that passes without action makes the disparity worse. Mandates without resources produce exactly what you would expect.