Building Your Team's AI Capability Roadmap
An AI capability roadmap is not a technology plan. It is a people plan that happens to involve technology. The question it answers is not "what tools will we deploy?" — it is "what should every person on my team be able to do with AI in 30, 90, and 180 days?"
Start by segmenting your team into three groups based on where they currently are:
- AI naive: They have limited or no practical experience with AI tools. Their primary need is foundational literacy and a safe place to experiment. Don't start them with complex enterprise tools — start them with consumer tools on non-sensitive tasks.
- AI curious: They have experimented with AI tools personally but have not integrated them into their work workflows. Their primary need is specific use case training and permission to apply what they know in their actual job.
- AI competent: They are already using AI tools regularly and effectively. Their primary need is advanced techniques, better tooling, and opportunities to help others develop. These are your change agents for the rest of the team.
Map your team against these three segments. You likely have a few AI competent, a larger number AI curious, and the remainder AI naive. Your roadmap addresses each group differently.
Change Management: Why People Resist AI and How to Address It
Resistance to AI is not irrational. It is a rational response to perceived threat and uncertainty. Your job as a manager is to diagnose the specific type of resistance so you can apply the right response — not to dismiss it or mandate through it.
Type 1
Fear of Job Loss
Signal: "Is this going to replace me?" — often unspoken but present in the subtext of resistance.
Response: Address it directly and honestly — not with platitudes, but with a concrete framing of how this tool changes the job rather than eliminates it. Show them a specific example where the AI handles the most tedious part of their role, freeing them for the work that actually requires their expertise. Then follow through — make sure that freed capacity goes to higher-value work, not just more of the same tedious work at higher volume.
Type 2
Trust Deficit
Signal: "I tried it and it was wrong." Usually preceded by one bad experience that has become a defining story.
Response: Validate the experience first — the AI was wrong, and that matters. Then reframe the role: the AI is a draft generator, not a decision maker. Show them specifically how to use the tool in a mode where errors are easy to catch and correct. Start with the lowest-stakes use cases where errors are obvious and low-cost. Build trust incrementally. Don't try to argue them out of their skepticism — earn their trust back one task at a time.
Type 3
Workflow Disruption
Signal: "I already have a system that works. Why should I change it?"
Response: The honest answer is: you shouldn't change for change's sake. Acknowledge that their current workflow works. Then make the specific case for why this change saves them time or effort they would rather spend elsewhere. Don't mandate adoption — invite it by removing friction. Make the new workflow easier than the old one, or the mandate will produce compliance theater rather than genuine use.
The most powerful change management tool: a respected peer who adopts voluntarily and shares their experience. Top-down AI mandates produce resentment. Peer modeling produces adoption. Identify your AI competent team members and give them visible opportunities to share what they've learned. Let their success do the selling.
What Good AI Training Looks Like
Most corporate AI training is terrible. It consists of a 60-minute slide deck about what AI is, followed by a generic demo of a consumer tool, followed by a quiz. Three weeks later, nobody has changed their behavior.
The training that actually changes behavior has three characteristics:
- Specific to the job. Show people how to use AI for the exact tasks they do every day — not generic use cases. "Here is how to use AI to draft the weekly status report" beats "here is how AI can help with writing."
- Hands-on during the session. Participants use the tool during training, not just watch a demo. The best training format is: 10 minutes of context, 20 minutes of guided practice on a real task, 10 minutes of debrief. Repeat.
- Followed by immediate application. Training that isn't applied within 48 hours is largely forgotten. Build training into a workflow so participants use the tool the same day or the next day on a real task.
The Manager's Evolving Role
The conventional management role — assign tasks, review outputs, escalate issues — is being compressed by AI. When AI can generate a first draft, analyze a dataset, or summarize a meeting in seconds, the "task assigner" part of management becomes less valuable. What becomes more valuable?
From task assigner to capability builder
Your primary job becomes building the capability of your team to work effectively with AI tools. That means evaluating tools (Day 2), building the business case (Day 3), running pilots that prove value (Day 4), and managing the human side of adoption (this lesson). The managers who thrive in the next decade are the ones who build AI-capable teams, not the ones who ignore AI until it's forced on them.
From reviewer to quality editor
When your team uses AI to generate first drafts, your review role changes. You are no longer reviewing from scratch — you are editing with a critical eye. The skill is knowing what to trust, what to verify, and what to rewrite. This is a different cognitive mode than traditional management review, and it is one worth practicing explicitly.
From knowledge holder to judgment applier
AI democratizes knowledge. The 15 years of institutional knowledge that made you valuable is now accessible to anyone who can query the right AI system. What remains irreplaceable is judgment — the ability to weigh competing considerations, navigate political complexity, make calls under uncertainty, and take responsibility for outcomes. That is not going away. Protect it by ensuring your team develops their own judgment, not just their ability to prompt AI.
Day 5 Exercise — Final Deliverable
Create a 90-Day AI Adoption Plan for Your Team
Using the template below, build the plan you will actually execute. This is not a presentation to leadership — it is an operational plan for you as a manager.
90-Day AI Adoption Plan Template
TEAM: [Your team name/size]
Manager: [Your name]
Start Date: [Date]
TEAM AI SEGMENTATION (as of today)
AI Competent: [Names/count] — Role in plan: change agents
AI Curious: [Names/count] — Role in plan: early adopters
AI Naive: [Names/count] — Role in plan: supported adoption
PHASE 1 — DAYS 1-30: Foundation
Goal: All team members have tried AI for at least one real task.
Actions:
Week 1: [2-hour training session — tool + specific use case]
Week 2: [Pilot with AI Competent + AI Curious members]
Week 3: [Peer sharing session — what worked, what didn't]
Week 4: [Structured reflection: baseline vs. results so far]
Success metric: [% of team that has used AI for at least 1 task]
PHASE 2 — DAYS 31-60: Integration
Goal: AI is part of the regular workflow for key tasks.
Actions:
[2-3 specific workflow integrations to implement]
[Any additional training needed — specific gaps identified in Phase 1]
[Tool or process changes based on Phase 1 learning]
Success metric: [Adoption rate for target tasks]
PHASE 3 — DAYS 61-90: Optimization
Goal: Measurable productivity improvement. Team advocates, not just users.
Actions:
[Advanced use cases for AI Competent members]
[Final measurement: time savings, quality, Team NPS]
[Lessons learned documentation for next initiative]
[Recommendation to leadership: expand, continue, or adjust]
Success metric: [Time savings + Team NPS targets]
CHANGE MANAGEMENT PLAN
Fear of replacement — I will address this by: [specific action]
Trust deficit — I will address this by: [specific action]
Workflow disruption — I will address this by: [specific action]
MY PERSONAL COMMITMENT
I will model AI adoption by: [specific thing you will do with AI
yourself, visibly, so your team sees you leading by example]
You have completed the course.
In 5 days you built: a task categorization framework, a tool evaluation scorecard, a real business case, a pilot design, and a 90-day adoption plan. That is a complete toolkit for leading AI adoption at your organization.
The next step is not more studying. It is implementing one thing from Day 1 of this plan.
Take the Next Step →
Course Summary: What You Now Know
- Day 1: AI is a pattern-completion engine. The 4 task types: generation, analysis, classification, conversation. Where AI excels and where it fails. The 5-question BS detector for vendor pitches.
- Day 2: The 5-dimension evaluation framework: security, accuracy, cost, integration, usability — with hard gates on security and accuracy. Three-phase evaluation process. Six vendor red flags. Build vs. buy decision matrix.
- Day 3: ROI calculation with conservative 50% time savings assumption. Risk assessment across four dimensions. The one-page business case format that gets approved. Pre-sell your skeptics. Start with a pilot.
- Day 4: Pilot selection criteria: small scope, high visibility, measurable. 30-day playbook week by week. Five metrics: time, quality, adoption, errors, Team NPS. Four failure modes to avoid.
- Day 5: Three-segment team capability roadmap. Three types of resistance and specific responses. What good AI training looks like. The manager's evolving role: capability builder, quality editor, judgment applier.
Accelerate your team's AI adoption
Our 3-day bootcamp covers advanced AI leadership, live tool evaluation, and hands-on strategy workshops — for you and your entire team. Five cities. $1,490 per seat.
Reserve Your Seat →