Introduction

‍

Your team is already using AI. Some people use it every day, some use it quietly, some avoid it entirely, and at least one person is probably pasting company information into a tool nobody approved. For managers in 2026, the question is no longer whether AI belongs in the workflow. It is whether you are leading how your team uses it, or letting it happen to them.
‍

The research on this is unusually clear. Gallup's 2026 workforce data shows that the strongest lever on AI adoption isn't the tool, the budget or the training portal. It is the manager. Yet as of early 2026, only around one in five employees strongly agreed that their manager actively supports their team's use of AI. That gap is where most of the value of AI is being lost.

This guide covers what has actually changed about managing a team in the age of AI, the four new responsibilities that now sit with every manager, the mistakes that quietly erode performance, and a 30-day plan to start leading an AI-ready team.

‍

Why leading a team using AI is now a core management skill

‍

For the last few years, most organisations treated AI as a technology rollout: buy the licences, run a webinar, measure logins.

The data now shows that approach stalls without managers at the centre.
‍

According to Gallup's AI indicator, employees who strongly agree that their manager actively supports AI use are 1.7 times as likely to use it regularly, 7.4 times as likely to say it gives them more opportunities to do what they do best, and 8.7 times as likely to say AI has transformed how work gets done in their organisation.

Gallup's mid-2026 engagement research found the same pattern: employees whose managers actively encourage and guide AI use reported engagement of 48%, compared with 30% for those without that support. When frequent AI use, a clear integration plan and active manager support all came together, engagement rose to 53%.

‍

BCG's global AI at Work survey tells a similar story. The share of employees who feel positive about generative AI jumps from 15% to 55% when they have strong leadership support, but only about a quarter of frontline employees say they actually receive it.

‍

The problem is that managers have rarely been prepared for this role. In a Gartner survey, just 8% of HR leaders said their managers had the skills to use AI effectively, and Gallup found that half of CHROs are not confident in their managers' ability to guide employees on AI. Organisations have invested heavily in the tools and very little in the people who decide whether those tools get used well.

‍

What has actually changed about the manager's job

‍

The administrative core is being automated
‍

Meeting notes, scheduling, status updates, progress tracking and first drafts of reports are increasingly handled by AI. This is the coordination layer that used to consume a large share of a manager's week. On paper, that frees time for higher-value leadership work. In practice, the time only gets reinvested if managers consciously redirect it towards coaching, problem-solving and building trust.

‍

Spans of control are getting wider
‍

Gallup reports that the average number of people reporting to a single manager rose to 12.1 in 2025, up from 10.9 the year before. Gartner has predicted that one in five organisations will use AI to flatten their structures and remove more than half of their middle-management roles. But Gartner also warns that by 2028, 75% of organisations that cut middle managers because of AI investments will end up hiring back at least half of those roles. Cutting layers without redesigning what managers actually do creates overloaded leaders, not efficient ones. Gartner's newer guidance suggests organisations shift from thinking about span of control to span of complexity: what a manager can realistically lead when both people and AI systems are in the mix.

‍

You now lead people and AI agents
‍

Microsoft's Work Trend Index introduced the idea of the agent boss: someone who builds, delegates to and manages AI agents alongside human colleagues. Microsoft found that 67% of managers were already comfortable with AI agents, compared with only 40% of employees. That gap matters. When managers race ahead of their teams, adoption becomes uneven, anxiety rises and the team splits into those who are confident with AI and those who quietly fall behind.

‍

The four new responsibilities of an AI-ready manager

‍

1. Set the rules before the tools
‍

When people don't have clear guidance or the right tools, they improvise. BCG found that more than half of employees without the AI tools they need will find alternatives and use them anyway, creating security and consistency risks. Meanwhile, Gallup found that only a quarter of US employees say their organisation has communicated a clear plan for integrating AI.
‍

You don't need to wait for a company-wide policy to give your team clarity. Agree a simple team AI charter that answers four questions:

• Which tools are approved for which types of work?
• What information must never go into an AI tool (client data, personal data, confidential strategy)?
• Where is human review mandatory before anything leaves the team?
• When should people disclose that AI was used, and to whom?
  ‍

2. Redesign workflows, not just tasks
‍

Most teams start by using AI to speed up individual tasks: a faster email, a quicker summary. The real gains come from rethinking how work flows through the team. At Asana, for example, the focus for managers has shifted from encouraging AI adoption to redesigning workflows with AI built in.

A practical way to start is to map one recurring team workflow and label every step as one of three things: AI drafts, human decides; human and AI collaborate; or human only. That exercise surfaces where AI genuinely helps, where it introduces risk and where human judgement is non-negotiable.
‍

3. Protect quality and stop workslop
‍

Researchers from Stanford and BetterUp Labs coined the term workslop for AI-generated content that looks polished but doesn't actually move the work forward. In their survey of US desk workers, around 40% had received workslop in the past month, and each instance took close to two hours to decode or redo. The interpersonal cost was even bigger: roughly half of respondents viewed colleagues who sent it as less capable and reliable, and 42% saw them as less trustworthy.

Workslop is a management problem before it is a technology problem. The researchers' recommendation is for leaders to model purposeful AI use, set clear norms and encourage a mindset that treats AI as a collaborator rather than a shortcut. The simplest norm to set: you own everything you send, whether AI helped or not.
‍

4. Coach the human skills AI can't replace
‍

The Harvard Business School and BCG field experiment with 758 consultants found that people using AI completed 12.2% more tasks, 25.1% faster and at higher quality, but only on tasks within AI's capabilities. On tasks outside that frontier, consultants who leaned on AI were more likely to get the answer wrong. The skill that separates the two outcomes is judgement: knowing when to trust the output, when to challenge it and when to put it aside.
‍

That makes critical thinking, context, ethical reasoning, communication and relationship-building more valuable, not less. The time AI frees up should flow directly into developing these skills through regular one-to-ones, feedback and stretch assignments.
‍

Common mistakes managers make when introducing AI to their team
‍

• Measuring usage instead of outcomes. Logins and prompt counts reward activity, not value. Teams learn to look busy with AI rather than work better with it.
• Using AI privately while asking the team to use it openly. Teams copy what managers do, not what they say. If you use AI, show how, including where it went wrong.
• Outsourcing the human moments. Performance conversations, difficult feedback and recognition should never feel machine-written. If your team would rather talk to a chatbot than to you, that is a trust signal worth acting on.
• Assuming junior staff will figure it out. Early-career employees often adopt AI fastest but have the least context to judge its output. As spans widen, protect time for mentoring rather than letting it disappear.
• Treating AI skills as a one-off training event. A single workshop rarely changes behaviour. New habits stick when people practise them in real work, reflect together and get feedback over several weeks.
  ‍

A 30-day plan for leading your team through AI adoption
‍

Week 1: Listen and experiment yourself. Ask each team member how they currently use AI, with no judgement. You will likely find a mix of power users, sceptics and quiet experimenters. At the same time, apply AI to your own workflow first so you can speak from experience.
‍

Week 2: Agree your team AI charter. Work through the four questions above together. Co-creating the rules builds ownership and psychological safety, and it surfaces concerns early.
‍

Week 3: Redesign one workflow. Pick a recurring process that frustrates the team, map it, decide where AI fits and run it as a pilot. Nominate a peer champion to support colleagues who are less confident.
‍

Week 4: Review, share and scale. Look at what improved, what created rework and what surprised people. Share both wins and failures openly, decide what to keep, and add a short AI retrospective to your regular team meetings.
‍

How to measure whether it's working
‍

Track a small set of signals that reflect real value rather than activity:
‍

• Frequency of meaningful use: are people using AI regularly within agreed workflows?
• Quality and rework: is the team spending less time fixing outputs, not more?
• Time reallocation: is freed-up time going into higher-value work, coaching and customers?
• Engagement and confidence: do people feel supported, or anxious and left behind?
• Business outcomes: faster cycle times, better client results, fewer errors.
  ‍

Building AI-ready managers across your organisation
‍

The evidence points in one direction: AI succeeds or stalls at the level of the manager. Yet most managers have never been trained to lead a hybrid team of people and AI, to set norms for responsible use or to coach the judgement skills that make AI valuable rather than risky.
‍

That is exactly why we built AI Leadership Labs at Affinity Space. They are applied Labs that help managers turn their team's AI use into better teamwork and real ROI delivered live by AI and leadership experts, designed to develop your organisation's new wave of AI-ready managers. Every participant leaves with practical toolkits they can use immediately, and our Affinity Missions make sure new skills are applied in the flow of work, not forgotten after the session ends.
‍

For broader leadership development, from new managers to senior leaders, explore our Leadership Labs. Our programmes are customised to your context, backed by neuroscience and measured for real learning transfer, with an average session NPS of 74.
‍

Ready to build managers who can lead the AI-enabled workplace? Book a call with our team.
‍

Frequently asked questions
‍

How do you lead a team that uses AI?
‍

Start by setting clear team norms for which tools to use, what data stays out and where human review is required. Then redesign workflows together, model transparent AI use yourself, hold everyone accountable for output quality, and invest the time AI saves into coaching and development.
‍

What skills do managers need to lead AI adoption?
‍

Beyond basic AI literacy, managers need change leadership, workflow design, the ability to coach critical thinking and judgement, clear communication about expectations, and the confidence to create psychological safety so people can experiment and admit mistakes.
‍

Will AI replace middle managers?
‍

AI is automating much of the administrative work managers used to do, and some organisations are flattening their structures as a result. But Gartner predicts that most organisations that cut middle managers because of AI will rehire many of those roles. The role is changing from supervising tasks to developing people and orchestrating work across humans and AI.
‍

How do you stop AI workslop in a team?
‍

Set a clear standard that people own everything they send regardless of how it was produced, agree where human review is mandatory, train people on when AI is and isn't appropriate, and model careful, purposeful use yourself.