How do I get partners at my firm to actually use AI?
It depends on why they are holding back. If they lack time, give them short, private, one-on-one sessions on their own work rather than group classes. If they doubt the quality, show AI on a task they know well so they can judge the output themselves. If they worry about looking unskilled, keep early practice private. If they simply see no reason, tie AI to something they care about, like client turnaround or capacity. Most partners respond to a useful result on their own work, not to mandates.
Who this is for
COOs, operations leads, heads of people and managing partners at professional firms where junior staff are trying AI but the senior people, whose example matters most, are not.
The short version
- Partner adoption matters because juniors copy what leaders do and what leaders reward.
- Senior people usually hold back for practical reasons: time, doubt about quality, or not wanting to look new at something.
- Private, short, hands-on sessions on their real work beat group training for this audience.
Why partner adoption matters so much
In most firms, partners set what counts as good work. If they never use AI, juniors learn that it is risky to use it openly, or that AI-assisted drafts will be judged harshly. Adoption stalls even when the tools are good. Our page on why AI adoption stalls covers this pattern in more detail.
Partners also hold much of the judgment and client context that makes AI output useful. When they use it, they find higher-value uses than juniors usually can.
Why senior people hold back
The reasons are usually reasonable, and each one needs a different response:
- No time. Partners’ calendars are full of client work and billable hours. A two-hour group workshop is an easy thing to skip.
- Doubt about quality. They have seen a weak AI draft once and concluded it isn’t good enough for their standards.
- Not wanting to look unskilled. Learning a new tool in front of juniors is uncomfortable for someone used to being the expert.
- No clear reason. Their current way of working got them to partner. Without a specific benefit, change feels optional.
- Concern about risk. Confidentiality, accuracy and professional liability are real concerns, especially for client-facing work.
Make sessions private and short
Offer one-on-one sessions of 30 to 45 minutes, scheduled around their calendar. Work in their own files and on their own tasks. Privacy lowers the cost of asking basic questions, and short sessions fit the way senior people actually spend time.
Start with their work, not features
Ask what they spent too long on last week. Common starting points for senior people include:
- Turning meeting notes into a client follow-up email.
- Getting a first read of a long document before a call.
- Drafting a memo or proposal outline from bullet points.
- Preparing questions for a client or counterparty meeting.
- Reviewing a junior’s draft for gaps or unclear points.
A useful result on a real task in the first session does more than any explanation of what AI can do.
Let them judge the quality
Senior people are good at spotting weak work. Use that. Have them run AI on a task they know deeply, then critique the result and improve the prompt. They see both the limits and the value, and their skepticism becomes a skill: checking AI output well. This is also what you want them to model for juniors.
Address risk directly
Explain which tools are approved, what data can go into them and how client information is protected. A clear AI acceptable use policy answers most of these questions. Partners are far more willing to try AI when they know they aren’t creating a confidentiality problem. See also how to keep client data safe when using AI.
Ask them to lead visibly
Once a partner finds one or two good uses, ask them to share it briefly in a team meeting or say openly that a draft started with AI. This signals to juniors that AI-assisted work is welcome, provided it is checked. It is often the single biggest lever for firm-wide adoption.
What doesn’t work
- Mandating usage targets for partners. It breeds resentment and box-ticking.
- Group sessions that mix partners and juniors. Senior people are less likely to ask questions or practice openly.
- Generic demos. Showing features unrelated to their work confirms the view that AI isn’t for them.
- Leaving it to enthusiasts. A keen junior offering to help is kind, but partners often won’t take it up.
How to know it is working
Track a few simple signals over two or three months. Are partners active in the approved tool? Do they mention AI in team meetings? Are juniors more open about using it? Are review cycles on drafts getting shorter? Our guide on measuring AI training results covers how to set this up without heavy reporting.
Common mistakes
- Treating partners as the last group to train. Start with them or alongside early adopters, because their example shapes everyone else.
- One session and done. A follow-up two weeks later, to fix what didn’t work, is where habits form.
- Ignoring the review role. Partners need to learn how to review AI-assisted work from juniors, not just how to produce their own.
When you do not need us
If you have someone internally who is trusted by senior people, understands the work and can give private sessions, you can run this yourself with the approach above. Outside help is most useful when there is no one senior people will learn from internally, or when you want a structured program across the whole firm.
How Ruby helps
- AI training: private, hands-on sessions for partners and senior leaders on their own work, plus role-based training for the rest of the team.
- AI strategy: aligning leadership on how AI should be used and reviewed across the firm.
- AI engineering: building tools for the repetitive work partners want off their desks, with them kept in charge of judgment and sign-off.
Many firms start with one practical hour of AI training for their senior team.