The truth about leadership, trust and AI adoption.
We asked 24 leaders and executive coaches one simple question: what separates organizations where AI adoption is working from where it is stalling? The answer is not the tools. It is the leaders.
AI arrived as a technology challenge but quickly manifested as a human one. People are being asked to change how they work, how they communicate, and in some cases how they understand their own professional identity, at a pace that makes thoughtful adaptation genuinely hard. Whether employees adapt or resist depends less on the tools available than on the actions of the leaders around them.
A semi-structured qualitative study using reflexive thematic analysis. We went deep with a small, deliberate sample of leaders who are living the AI transition in real time.
Twenty organizational leaders across HR, L&D, change, comms, product, technology, sales. Plus four executive coaches with broad visibility into C-suite behavior.
Each interview audio-recorded and transcribed verbatim. Three domains: personal AI use, AI's impact on leadership effectiveness, and driving AI adoption in others.
Hierarchical coding with up to three levels per snippet. Two frequency metrics tracked: snippet count and participant breadth, so depth and reach both stay visible.
Twelve signals from the dataset organized by theme. The percentages reflect the share of our 24 participants who raised each idea.
Every interviewee reported using AI in their work. This is not a study of adopters versus skeptics. It is a study of how thoughtfully people use it.
Substantial, but not yet a majority. A further 17% described AI as integrated into regular workflow but not quite daily.
The leaders with the best outcomes are not the heaviest AI users. They are the most intentional ones. Ubiquity without intentionality is a liability, not an asset.
Communication is the leadership capability most affected by AI, with problem-solving close behind. Innovation and creativity follow at 21%.
People who figure out how to use AI for problem-solving will go further. AI lets leaders focus on higher-value thinking instead of executional drag.
Human relationships, judgment, emotional presence, and the ability to make meaning from ambiguous information remain distinctly and irreplaceably human.
The dominant risk to leadership credibility is not using AI. It is distributing obvious, unedited AI output. The corrective is to edit ruthlessly and own what you put your name on.
Authenticity concerns are mounting. Several participants described identity threat: a core professional self-concept challenged by a tool.
Leaders gained credibility by showing up better prepared, with deeper insights and confident data use. AI as research accelerator, not as a ghostwriter.
When leaders use AI openly, reference it in conversations, and share both wins and failures, teams follow. When leaders are absent, adoption stalls.
Especially fear of replacement. The most effective antidote is naming the anxiety directly, not minimizing it. Cost-savings framing validates the fear.
The single most prevalent code in the entire context section. Global matrices, multi-layered hierarchies, and influence without formal authority.
These are the headline numbers.
The full report covers 249 distinct codes across 1,190 participant snippets, with every quote, the complete methodology, and the practitioner playbook.
Download the Full ReportClick any card to read the finding, the supporting evidence, and the Monday Move you can run with this week.
When leaders use AI openly, reference it in conversations, and share both wins and failures, teams follow. When leaders are absent from the conversation, adoption stalls. Modeling was named the single most powerful adoption behavior across 83% of participants. This is not close to second.
Coach one leader to share an AI experiment in their next team meeting, including what did not work. Vulnerability builds more trust than polish.
The dominant risk to leadership credibility is not using AI. It is distributing obvious, unedited AI output. 58% of participants recognized this pattern with pattern-recognition fluency. Another 29% worry their voice will be "blanded" by AI. The corrective is not to avoid AI. It is to edit ruthlessly, maintain your voice, and never distribute something you cannot stand behind and explain.
Pick one AI-assisted message this week. Read it out loud before you send. Does it still sound like you? If your name is on it, you own it.
Fear, especially fear of replacement, is the dominant emotional barrier to adoption. 83% of participants named it. The most effective antidote is not minimizing the fear but addressing it directly: name the anxiety, separate augmentation from replacement in your language, and demonstrate through action that AI is making the work better rather than making the worker obsolete. When organizations frame AI primarily in terms of cost efficiency and headcount reduction, they validate the replacement fear.
Audit your AI talk track. Does it lead with cost-savings or with capability? Cost-saving framing validates replacement fear. Capability framing invites partnership.
The leaders who describe the best outcomes are not the heaviest users. They are the most intentional ones. The mindset shift matters more than the tool: thinking of AI as a thought partner rather than a shortcut, a first drafter rather than a ghostwriter, a collaborator rather than a replacement. Ubiquity without intentionality is a liability, not an asset.
Before opening your AI tool today, finish this sentence: "The human judgment I'm bringing to this is ___." If you can't answer, do the human work first.
Human relationships, judgment, emotional presence, and the ability to make meaning from ambiguous information remain irreplaceably human. As AI raises the floor on analytical, drafting, and research tasks, distinctly human capabilities become the differentiator. The more AI handles, the more human the leader must become. Communication is the leadership capability most enhanced by AI (54%), but only when paired with genuine voice and human judgment.
This week, name one task you have been delegating to AI that should not be. Take it back. Notice what shows up when you do the human thinking yourself.
15 questions. Five minutes. A personalized AI Leadership Credibility profile mapped to the five dimensions of our research. You leave with three Monday Moves you can run this week.
Answer honestly. The whole point is to find your real growth edges, not score well. Research participants showed a consistent gap between self-rating and team-rating, and the gap was largest on the dimensions most visible to followers.
Mapped to the five dimensions from From Hype to Human.
Three actions from your two lowest dimensions. Pick one, run it this week.
If you advise leaders through transformation, these are the highest-leverage interventions from our research. Want more tailored tips? Take the self-assessment to identify your specific gaps and get personalized recommendations.
Help leaders narrate their AI experiments in meetings, including the failures. Vulnerability builds more trust than polish (38% of leaders said transparency about wins and failures builds more trust than polished success stories).
Cost-cutting framing validates replacement fear. Connect AI use to integrity, growth and the work people actually care about. The WIIFM principle showed up across this research as the difference between adoption and resistance.
Experimentation and piloting was the single most-cited enablement approach (83% of participants). Create explicit permission to try, fail and share what worked. People who try AI in low-stakes contexts discover value, and discovery creates momentum.
Full methodology, all five findings unpacked, the complete coding scheme, dozens of unedited participant quotes, and a practitioner playbook with the frameworks behind every Monday Move.
Enter your email to download the full report:
Findings citable with attribution. No spam, just occasional research updates.


