How to Communicate AI-Assisted Work to Stakeholders
As AI becomes a standard part of professional work, your ability to explain how you use it determines whether stakeholders trust or question your output. This playbook gives you concrete techniques for describing AI's contribution honestly, setting realistic expectations, and handling the misconceptions that derail productive conversations. The goal is transparency that builds confidence, not hype that creates inflated expectations or secrecy that breeds suspicion.
This playbook covers the how. For the why and what, see the
skill definition
.
Developing Start here. Build the foundation.
- Write a 3-sentence standard explanation of how you use AI in your most common deliverable. Follow this structure: (1) what the AI does ('I use AI to generate a first draft of the weekly summary from raw notes'), (2) what you do ('I then verify every data point, adjust the framing for our audience, and add strategic context'), (3) why this approach works ('This cuts drafting time in half while keeping accuracy at the same level as my fully manual process'). Practice saying it out loud until it feels natural, then use it the next time someone asks.
- Create a simple one-page reference card with two columns: 'What AI can do well' and 'What AI cannot do well' for your specific work context. For example: 'Can do well: summarize meeting notes, draft routine emails, organize data into tables. Cannot do well: verify factual accuracy, understand internal politics, make judgment calls about client relationships.' Share this with your direct stakeholders in your next check-in and ask if they have questions.
- The next time you present AI-assisted work, add one sentence of disclosure before diving into the content. Use this format: 'I used [tool name] to [specific task], then [what you did to validate and improve it].' Keep it matter-of-fact: no apologizing and no overselling. Notice how people react and adjust your level of detail based on whether they want more or less information.
Proficient Build consistency and rhythm.
- Build three versions of your AI explanation calibrated to different audiences. Version 1 (Executive): Focus on business impact: time saved, quality maintained, risk managed. Version 2 (Technical peer): Focus on process: tools used, prompts that work, validation steps. Version 3 (Non-technical stakeholder): Focus on outcomes: what they receive, how it was checked, what to flag if something looks wrong. Keep all three in a note you can reference before any meeting where AI use might come up.
- When someone makes an incorrect assumption about AI (either overestimating or underestimating its capabilities), use the 'Yes, and actually' technique. Acknowledge their perspective, then redirect with a specific example: 'You are right that AI can draft emails quickly. And what it actually needs from me is the strategic framing and the decision about what to emphasize for this particular client.' Prepare 3-4 of these redirects for the misconceptions you hear most often.
- Add a standing agenda item to your next team or stakeholder meeting: '2-minute AI transparency update.' Use it to share one specific example of how AI helped that week, including what worked, what you had to fix, and what you learned. This normalizes the conversation and prevents AI use from feeling like a secret. Run this for 6 consecutive meetings, then evaluate whether to continue or reduce frequency.
Mastered Operate at the highest level.
- Draft a disclosure protocol for your team or workgroup that answers four questions: (1) When should we disclose AI assistance? (2) How much detail should we provide? (3) Where should we document it? (4) Who decides in ambiguous cases? Keep it to one page. Circulate it for feedback, finalize within 2 weeks, and review it quarterly. This removes guesswork and creates consistency across the team.
- Run a quarterly 'stakeholder perception check' by asking 3-5 key stakeholders two questions: 'Do you feel you understand how our team uses AI?' and 'Is there anything about our AI use that concerns you or that you would like to know more about?' Compile the responses, identify gaps in your communication, and adjust your messaging. If the same concern appears twice, address it proactively in your next update.
- Develop a 15-minute presentation you can deliver to any new stakeholder group that covers: what AI tools your team uses, what tasks they handle, what your validation process looks like, and what the realistic benefits and limitations are. Include 2-3 concrete before-and-after examples showing AI-assisted vs. manual output. Deliver it at least once per quarter to new partners, clients, or cross-functional teams you start working with.
Unlock Skill Progression
Coaching Personalized to your current level
Progress Tracking Across every skill area
Mastery Validation Evidence-based, not guesswork