Every firm I talk to uses AI tools by now, whether they’re built into their standard tech stack or a general-purpose platform like Claude, ChatGPT, or Microsoft Copilot. Few firms have an AI-first culture, and it explains why some firms are seeing real productivity gains from AI while others are paying for compute credits they’re not sure what to do with.
Buying a tool is a procurement decision. Building a culture is a behavioral one. An AI-first culture puts thought into the expectations and guardrails you put around how people think and work, day to day, no matter which platform you license or how many team members logged in this month. You can hand every team member the same AI assistant and get wildly different results, because the habits (not the tool) determine the outcome.
That understanding is essential right now because trust in AI at work is still shaky, and leaders and team members aren’t seeing it the same way. RSM’s 2026 Middle Market AI Survey found that 85% of senior leaders agreed their organization’s executive leadership is more enthusiastic about AI than employees are. That disconnect shows up as hesitation, workarounds and resistance long before it shows up in performance reviews. You won’t close that gap by ignoring it. You must close it on purpose by modeling and reinforcing the behavior you want to see.
AI-first doesn’t mean people last
Before we get into what to build, let’s clear away what an AI-first culture is not, because the misconceptions are damaging.
AI-first doesn’t mean replacing people. If your team members hear “AI-first” and translate it to “job cuts,” you’ve already lost the room, and you’ll spend the next year fighting adoption instead of building it.
AI-first also doesn’t mean automating everything. Some of your firm’s value lives in the judgment calls, client relationships and professional skepticism that no model can replicate.
AI-first doesn’t mean blind trust. In an AI-first firm, team members don’t hand off their thinking to a chatbot and accept whatever comes back. That’s outsourcing your judgment, and it’s how “AI slop,” the vague, generic, unreviewed output that erodes client confidence, ends up in a deliverable with your firm’s name on it.
What AI-first actually means
AI-first means AI becomes part of how someone thinks through a problem before they escalate it. It’s a discipline, not a tool subscription. The behavior you’re trying to instill is simple to describe and hard to sustain. Before a team member brings a problem to a manager, schedules a meeting or drafts an email asking someone else to do something, they pause and ask themselves one question.
“How would AI help me think through this first?”
That one question, when asked consistently, changes how work moves through a firm. AI won’t solve every problem, but every problem gets a first pass of independent thinking before it becomes someone else’s problem.
Build these behaviors for an AI-first firm
If you’re trying to make this concrete for your team members, here are a few specific habits to nurture:
- Prompt before pinging. Before firing off a message to a colleague asking “how do I…” or “can you send me…,” try getting a first answer from AI. Save the ping for what AI genuinely can’t resolve.
- Draft before delegating. Before handing a task to someone else, generate a rough first draft yourself. Even an imperfect draft gives the person you’re delegating to something to react to, which is faster than starting from a blank page.
- Summarize before escalating. Before bringing an issue up the chain, use AI to help you organize what’s going on. Explain what the issue is, what you’ve tried and what decision you need. Managers spend less time untangling the problem and more time solving it.
- Analyze before meeting. Before you call a meeting to “figure something out,” see how far AI can get you on the analysis first. Bring options and a recommendation instead of an open-ended question.
These behaviors don’t ask anyone to unquestioningly trust AI. Each one keeps a person in the loop, reviewing, editing and deciding. Those guardrails keep output from turning into AI slop.
Earn trust in both directions
Research highlighted in Chicago Booth Review points to a trust deficit leaders need to take seriously. Only about one in five employees trust their leadership, according to Gallup, and separate research cited in the same piece found that leaders consistently rate themselves higher on accountability than their own direct reports rate them.
If your team members don’t see leaders modeling these behaviors themselves, no amount of policy language will make the culture stick. Leaders who draft their own first pass and summarize before escalating to the board or partner group make “AI-first” credible instead of aspirational.
Where this leaves your firm
The firms realizing value from AI right now might not have the most sophisticated tool stack. But their team members have become orchestrators of their own work. They use AI to think faster and then bring their best thinking forward. They don’t just hand work off to an AI tool because this isn’t a technology rollout. It’s a culture you build one interaction at a time, starting with how you, as a leader, work.
===
Jim Boomer is the CEO of Boomer Consulting, Inc.
Sign in to get access to this free resource, and all of our whitepapers and reports.
Download this content today!
Register Now Already registered? Click here to Log In
