Leading AI-Enabled Teams: How to Manage People Working Alongside AI
Deliberate Academy Editorial Team
Reviewed for accuracy and professional relevance
You're 9 lessons in — don't lose your progress.
Sign up free to save where you are and earn a verified certificate when you pass.
- Explain why team anxiety about AI is legitimate and describe what genuine leadership engagement with that anxiety looks like in practice
- Identify how human roles shift in an AI-enabled team and apply this to updating role descriptions in your context
- Apply the performance management shifts required in AI-enabled teams, including evaluating AI use as a professional skill
- Recognize the management actions required to build team-wide AI capability, rather than relying on individual self-directed development
You have deployed AI tools that have measurably improved output quality and reduced time on routine tasks. The tools are working. But your team is anxious. Your highest performers are worried about what AI means for their careers. Two team members are using AI enthusiastically but not disclosing it, creating questions about whose work is whose. One person refuses to use AI at all, citing concerns about accuracy. And you are not sure how to evaluate performance when some of the work is AI-assisted and some is not. The technology deployed successfully. The people management challenge is just beginning.
The Anxiety Is Real and Legitimate
Before any other management consideration, leaders of AI-enabled teams must genuinely engage with the anxiety that AI creates for their people. This anxiety is not irrational. The evidence that AI will change the nature of many jobs is real, even if the timeline and magnitude of change is uncertain. Dismissing the anxiety with reassurance that "jobs are safe" is both unconvincing and potentially dishonest. Engaging with it honestly is the only credible leadership approach.
What genuine engagement looks like: acknowledging that AI is changing the nature of work, being transparent about what you know and do not know about how it will affect your team's roles, being specific about which aspects of their work AI will handle versus which remain clearly human, and creating a forum where concerns can be raised without penalty.
Leaders who model honest engagement build the trust that makes AI adoption faster and more successful. Leaders who dismiss anxiety create an underground of fear and resistance that undermines adoption regardless of how good the tools are.
Have a direct conversation with your highest performers about AI and career development. These are the people with the most to potentially feel threatened by AI and the most options if they decide to leave. A proactive conversation about how AI changes their role, how you see their value evolving, and what development you will invest in is far more effective than hoping they reach their own conclusions.
Redefining Roles in an AI-Enabled Team
As AI takes on more of the execution of routine tasks — drafting, data processing, summarisation, formatting — the human role in those tasks shifts toward direction, judgment, and quality assurance. This is a genuine role change, not just a productivity increase.
The tasks that remain firmly human: setting the direction and standards that AI operates within, exercising judgment on outputs that involve significant ambiguity, maintaining the client and stakeholder relationships that depend on human trust and context, making decisions with ethical or political implications, and providing the creative insight or organizational knowledge that AI cannot access.
For team managers, this means role descriptions may need updating to reflect the changed balance of execution versus judgment. A content strategist in an AI-enabled team is doing more strategic direction of AI output and less drafting. A data analyst is doing more interpretation and less cleaning and formatting. If role descriptions still describe the execution-heavy version of the role, they are out of date.
Managing Performance in an AI-Enabled Context
Performance management in AI-enabled teams requires rethinking what you are measuring and what signals indicate high versus low performance.
Shift from output volume to output quality and judgment. If AI is handling the routine execution, volume alone is no longer a meaningful performance signal. What matters is the quality of direction given to AI, the judgment applied to AI outputs, and the value added in the human-owned parts of the work. Managers who still evaluate primarily on volume will inadvertently create incentives to use AI to inflate output volume without adding judgment value.
Address AI disclosure explicitly. Set a clear team norm on AI disclosure. Some organizations require employees to disclose AI use in submitted work. Others do not require disclosure but set quality standards that work must meet regardless of how it was produced. Either approach is defensible — inconsistency is not. The norm needs to be stated, not assumed.
Evaluate AI use as a skill. Using AI tools effectively is a genuine professional skill. The ability to write effective prompts, critically evaluate AI outputs, identify hallucinations, and integrate AI into a high-quality workflow is part of what a competent professional in 2026 looks like. Consider whether your performance framework recognizes and rewards this skill appropriately.
When a team member consistently produces AI-assisted work that is of high quality, demonstrates good judgment in directing AI, and clearly applies their own expertise in the non-automatable parts — that is strong performance. When a team member produces AI-assisted work without adequate review or quality control, that is a performance issue regardless of volume. The standard is the quality of the final output and the judgment applied, not the tool used to produce it.
A senior analyst on your team has doubled their output volume since adopting AI tools. However, you notice that several of their recent reports contain conclusions that are well-supported by the AI-generated analysis but lack the strategic interpretation and sector context that previously distinguished their work. How should you interpret this performance signal?
Select one answer.
Building Team Capability for AI-Enabled Work
You cannot rely on individuals to develop effective AI skills through unsupported trial and error. The team capability gap is real, and closing it is a management responsibility.
Effective team capability building in AI includes: structured training on the tools the team is expected to use, not just access to the tools; clear quality standards for AI-assisted work that people can calibrate against; a team prompt library or knowledge base so that effective approaches are shared rather than siloed; regular retrospectives where teams share what is working, what is not, and what they have learned; and recognition of effective AI use in team forums so that it is visible as a valued skill rather than a hidden shortcut.
The team members who develop the strongest AI capabilities first are your most valuable internal champions. Invest in their capability development, give them visibility, and leverage them to accelerate the development of the rest of the team.
Leading Through Ongoing Change
AI capabilities will continue to evolve, which means the management challenge of AI-enabled teams is ongoing, not a one-time transition. The leaders who manage this well share several characteristics.
They stay current with AI developments that are relevant to their team's work. They create regular opportunities for their teams to experiment with new AI capabilities without the pressure of immediate production use. They are honest when tools are not delivering expected value and willing to change course. They maintain the focus on the human dimensions of their team's work — development, relationships, culture, judgment — that AI cannot replace.
Redefining Performance Standards After AI Deployment — Marketing Agency
Context
A Head of Content Strategy deployed AI writing tools across a 12-person content team following a successful pilot. Output volume increased substantially in the first month. However, she noticed that several team members were producing more content but applying less strategic direction — accepting AI-generated angles without editorial judgment and reducing the distinctive positioning that had previously differentiated the agency's work.
Action
She recognized that the performance framework still evaluated primarily on volume and delivery speed, which had inadvertently incentivized AI volume inflation rather than quality elevation. She revised the team's performance criteria to assess three dimensions: quality of creative direction given to AI tools (evaluated through prompt review and brief-setting), judgment applied in editing AI outputs (assessed against client brief alignment), and the proportion of distinctively strategic work in each person's output. She introduced a monthly portfolio review where work was assessed on quality signals, not volume.
Outcome
Within two quarters, the team's output mix had shifted toward higher-quality strategic content with AI handling the execution. Two senior team members who had been coasting on volume metrics raised their performance substantially once the quality and judgment dimensions were measured. The agency attributed a renewal from its largest client partly to a noticeable improvement in strategic content quality during this period.
Why should performance evaluation in AI-enabled teams shift from output volume to output quality and judgment?
Select one answer.
Exercise
Your Task
Select one role on your team that is actively using AI tools. Produce a role redefinition document with three sections: (1) identify two or three existing role description bullet points that describe execution tasks AI is now handling or significantly supporting — quote them as-is; (2) rewrite each bullet to describe what the role now requires: the direction, judgment, quality assurance, or expertise the person applies rather than the execution they previously performed; and (3) write one paragraph describing what strong performance looks like in this role in an AI-enabled context — specifically naming the judgment and quality standards you would use to evaluate it.
Success looks like
- Each rewritten bullet describes direction or judgment activity, not task execution — 'reviews AI-generated analysis for strategic coherence and sector relevance' rather than 'produces competitive analysis reports'
- The performance paragraph names specific quality signals you would actually look for, not generic statements about 'effective AI use'
- The redefined role description would read as genuinely accurate to someone currently in that role — not aspirational or defensive
Watch out for
- Rewriting role descriptions to sound more impressive without changing what is actually evaluated — performance management must follow the updated description or the exercise is purely cosmetic
- Omitting the execution tasks that AI still requires human oversight for — roles in AI-enabled teams still have oversight and quality assurance responsibilities that must be named explicitly
Hint
Ask the person in the role what they spend most of their time on now versus six months ago — the gap between that answer and the current role description is where the rewriting should focus.
Try It: AI-Graded Practice
The exercise below grades your rewritten bullet automatically, checking whether it genuinely shifts toward direction and judgment rather than just sounding more impressive.
- The anxiety AI creates in teams is legitimate and requires honest engagement, not dismissal — transparency about what you know and do not know builds the trust that makes adoption faster and more successful.
- As AI handles more execution, human roles shift toward direction, judgment, and quality assurance — role descriptions should be updated to reflect this genuinely changed balance rather than describing the execution-heavy prior version.
- Performance management must shift from output volume to output quality and judgment — volume-based evaluation creates incentives to inflate AI output without adding the judgment value that makes the work professionally sound.
- Set explicit team norms on AI disclosure — inconsistency on this creates confusion about standards, accountability, and whose work is whose.
- Team capability building in AI is a management responsibility, not a self-directed individual exercise — structured training, shared prompt libraries, and public recognition of effective AI use accelerate team-wide capability development.