How do you lead a team that works with AI every day?
Managers are asked to get their teams using AI, and they answer for what goes wrong. The documented failures include unchecked output and untested workload claims. In May 2025 the Chicago Sun-Times printed a syndicated summer reading list in which 10 of 15 books did not exist; the writer said he used AI and did not check. Commonwealth Bank of Australia planned to replace 45 customer service roles with an AI voice bot in July 2025, then reversed the decision in August and apologised.
This course gives line managers one plan for their own team: which tasks AI does, assists or stays away from, who checks what, and how to tell whether it worked. You build the plan on your team’s tasks in one day, with a review date to test it against results.
After this course
- Can you sort your team's tasks into 3 groups: AI does it, AI assists, or AI stays away?
- Can you set a check routine that names who verifies which AI output, against which source?
- Can you write a quality standard and a sign-off line for AI-assisted work that leaves the team?
- Can you keep a person responsible for hiring, rating and other people decisions, with a record of each?
- Can you test a workload claim against your team's own data before roles or headcount change?
- Can you talk with staff about job worries using facts, open questions and dates?
Skills you practise
- Task assessment
- Check routines
- Quality standards
- Fair people decisions
- Workload testing
- Team measures
- Staff conversations
Course structure
The course is 1 day of 6 teaching hours, in 2 blocks of 3 hours with a lunch break. Online, it runs as 2 live sessions of 3 hours on the same day, not recorded.
The morning is about the work. You sort your team’s tasks into 3 groups and test the sort on examples from your own team. You then set the check routine: who verifies which output, against which source, and how findings are logged. The morning ends with quality standards for deliverables. In October 2025 Deloitte Australia refunded part of its fee for a government report that contained a made-up court quote and references to papers that do not exist.
The afternoon is about people. It starts with hiring, rating and other people decisions. Amazon dropped an experimental CV-scoring tool, reported in 2018, after it learned to penalise CVs that contained the word “women’s”. In EEOC v. iTutorGroup, settled in 2023 for USD 365,000, automated hiring software had rejected women aged 55 or over and men aged 60 or over. Mobley v. Workday is still before the court, and its claims that screening software discriminated by age, race and disability remain allegations. You also cover GDPR Article 22 on decisions made only by automated means.
Next come workload claims. In May 2025 Klarna’s chief executive said its AI cost-cutting push in customer service had gone too far and that the company would recruit human agents again. You test one claim against your own data and set 3 to 5 measures with baselines. You then prepare the first team conversation about job worries, with the facts you can share and dates for the answers you do not have yet. The last 45 minutes turn the work into a one-page team adoption plan with a review date.
About two thirds of the time is hands-on, in your organisation’s approved AI tool.
1Sorting tasks: AI does, AI assists, AI stays away75 min
- Listing the team's recurring tasks
- Tasks AI does and a person checks
- Tasks where AI assists a person
- Tasks AI stays away from
- Testing the sort on examples
Each task on the list gets one of 3 labels: AI does it and a person checks, AI assists a person, or AI stays away. The label sets how far the tool may go on that task.
Caution. Put tasks that involve personal or confidential data in the 'stays away' group unless your approved tool list allows that data class.
You leave with: A first task map for the participant's own team, with a test for each 'AI does' task.
2The team check routine60 min
- Who checks which output
- Checking against a source outside the chat
- Small steps that are quick to check
- Logging what the checks find
Every AI draft comes with a numbered list of its claims and their sources. The checker works through that list against the originals before anyone uses the draft.
Caution. A 2024 study found AI models rate their own writing higher than writing that human reviewers judged equal, so the checker should be a person or a different tool.
You leave with: A check routine for the team: which outputs get checked, by whom, against which source, and where findings are logged.
3Quality standards against AI slop60 min
- What counts as AI slop in a deliverable
- Made-up quotes and references
- Bought-in and outsourced content
- Telling clients when AI was used
- A sign-off line per deliverable
A model can give a confident answer that is wrong, including quotes and references that do not exist. A quality standard names what gets checked before anything leaves the team.
Caution. In October 2025 Deloitte Australia refunded part of the fee for a government report worth AUD 440,000 that contained a made-up court quote and references to papers that do not exist.
You leave with: A one-page quality standard for AI-assisted deliverables, with a sign-off line.
4People decisions and AI45 min
- Hiring, rating and dismissal
- GDPR Article 22 and human review
- Bias learned from past decisions
- Vendor tools and who carries the risk
- AI Act rules for employment uses from 2 December 2027
A model produces its output from patterns in the data it was trained on. If past decisions were skewed, the patterns carry the same skew.
Caution. Amazon dropped an experimental CV-scoring tool, reported in October 2018, after it learned to penalise CVs that contained the word 'women's'; keep a person responsible for every people decision.
You leave with: A team rule on which people decisions AI may support, and which records to keep.
5Workload claims and measuring results45 min
- Testing a time-saving claim on your own data
- Baselines before the change
- Quality and customer measures next to cost
- 3 to 5 measures per team
A time saving measured on one task says little about the next. A team-wide saving stays a guess until each task has been measured against a baseline.
Caution. In August 2025 Commonwealth Bank of Australia reversed its plan to replace 45 customer service roles with an AI voice bot and apologised, after the union said call volumes were rising.
You leave with: 3 to 5 team measures with baselines, and a test plan for the largest workload claim.
6Talking with staff about job worries30 min
- What you know and what is still undecided
- Sharing the task map with the team
- Questions staff ask and plain answers
- When to involve HR and employee representatives
Say which tasks AI will support and which, if any, it will take over. Tell the team task by task, including what is still undecided.
Caution. Since 2 February 2025 the AI Act has banned AI that infers employees' emotions at work, except for medical or safety reasons. Keep such tools out of any attempt to gauge how staff feel.
You leave with: Talking points for the first team conversation about AI, with dates for the open answers.
7Writing the team adoption plan45 min
- Task map, check routine and measures on one page
- Owners and dates
- What to tell the team and when
- Setting the review date
A long chat that mixes earlier drafts and side questions gives weaker answers. Start the plan in a new chat, with the task map, checks and measures in the first message.
Caution. Research published in 2024 found models miss information placed in the middle of long inputs, so put the key facts at the top of the brief.
You leave with: The team adoption plan, ready to share with the team and the participant's own manager.
Sorting tasks: AI does, AI assists, AI stays away
Give the task list and the criteria, then ask for a table with reasons.
Which of my team's jobs can AI do?
Below are 20 recurring tasks from my accounts payable team of 9. Sort each into one of 3 groups: AI does it and a person checks, AI assists a person, or AI stays away. Criteria: cost of an error, data sensitivity, and how easy the result is to check. Give a table with task, group, the criterion that decided it, and one test I can run this week to confirm the choice. [task list]
The team check routine
Ask for claims as a numbered list the checker can tick off.
Check this report for mistakes.
List every factual claim, number and quote in the draft below as a numbered line. For each, give the source the draft relies on and the page or cell. Mark any claim without a source as 'unsupported'. A colleague will check each line against the original. [draft]
Quality standards against AI slop
Ask for sources you can open, and 'not found' where none exists.
Add some references to make this report stronger.
For each claim in the report section below, give the source it rests on, with a link I can open. Where no source supports a claim, write 'not found' next to it. List only sources you can link to; I will open every one. [report section]
People decisions and AI
Ask for evidence per criterion, and keep the decision with a person.
Rank these 40 CVs and tell me who to reject.
For the role profile below, list the 6 job criteria in a table. For each anonymised CV, quote the lines that show evidence for each criterion and write 'no evidence found' where there is none. Stop after the evidence table; the hiring manager reads it and decides. [role profile] [anonymised CVs]
Workload claims and measuring results
Ask what data would prove the claim wrong.
AI will save my team 20% of their time, right?
A plan says an AI assistant will cut my team's time on supplier queries by [claimed saving]. List the data I need to test that claim: the baseline, the period, the quality checks, and the result that would show the claim is wrong. Then draft a 4-week test plan as a table. [plan extract]
Talking with staff about job worries
Give the facts you can share, and ask for plain answers to hard questions.
Write a message reassuring my team that AI won't replace them.
Draft talking points for a team meeting about AI. Facts I can share: the task map below, the 3 tasks we will test first, and that no role changes have been decided. List the 8 questions staff are most likely to ask, with an answer of under 40 words for each. Where I do not know the answer yet, say so and give the date we will know. Tone: direct and calm. [task map]
Writing the team adoption plan
Put every requirement in the first prompt, key facts at the top.
Help me write a plan. [later] Add measures. [later] Make it one page.
Write a one-page team adoption plan from the 4 inputs below: task map, check routine, measures, review date. Sections in this order: purpose (2 lines), tasks by group, who checks what, measures with baselines, owners and dates. Under 400 words, with headings and bullets. [inputs]
You leave with a team adoption plan on one page: the task map, the check routine, 3 to 5 measures with baselines and owners, and a review date. You also leave with a quality standard for deliverables and talking points for the first team meeting.
In the week after the course, you share the task map with your team and start the check routine on the first 3 tasks. The measures start counting from the baselines. Each participant receives an attendance record for your AI literacy file.
- Teaching time
- 6 hours on 1 day, in 2 blocks of 3 hours with a lunch break.
- Before the day
- None; participants bring a list of the tasks their team does in a typical month and, if one exists, their organisation's AI use policy.
- After the day
- A team adoption plan: a task map, a check routine, 3 to 5 measures and a review date.
Completion
You complete the course by working through each exercise on your own tasks. There is no exam and no grade. Each participant receives an attendance record for your AI literacy file (not a certificate).
Who is this course for?
Line managers, team leads and heads of department whose teams already use AI or are about to get an approved tool. It fits teams that send deliverables to clients, handle personal data, or have been told AI will save them time.
No technical background is needed. Bring a list of the tasks your team does in a typical month.
Choose Leading AI Adoption: Governance, Policy and Culture if your organisation has no AI use policy or approved tool list yet; this course works inside those rules. Choose AI for Daily Work: Prompts, Assistants and Automations if your team members need hands-on skills with assistants and automations.
In the Manager path, this course is step 3 of 3, after Leading AI Adoption.
- Level
- Intermediate
- Before you start
- None; participants bring a list of the tasks their team does in a typical month and, if one exists, their organisation's AI use policy.
- Group
- 1 to 12 people, online, live or at your office
This course is right for you if you
- manage a team that uses AI or is about to
- answer for the quality of what your team delivers
- have been told AI will save your team time
Unsure whether it fits? Ask us. We reply within 1 working day.
Practical information
- 1 day, 6 teaching hours, with a lunch break
- Online: live in 2 sessions of 3 hours on one day, not recorded; times agreed in the intake call
- At your office: 09:30 to 16:30; you provide a room, a screen and internet access
- In English, in the AI tool your organisation has approved
- Taught live by a practitioner with more than 15 years of commercial and technical leadership in software and data companies across Europe and Asia-Pacific, including maritime, energy and fintech.
- Attendance record for your AI literacy file
- Online, live: 1 day in 2 live sessions of 3 hours, not recorded. Book it as a private day for your team of 1 to 12 people, on your own Teams, Zoom or Meet.
- At your office: 1 day (6 teaching hours) for 1 to 12 people, travel within the Netherlands included.
Dates and booking
Choose the format, an open weekday in the calendar and the number of participants, then reserve on this page.
The calendar and booking form load with JavaScript. Without it, ask us for a date on WhatsApp, by phone on +31 6 20 51 73 96 or at advisory@ai-business.school.
Investment
| Item | Online, live | At your office |
|---|---|---|
| Price | €399 per person | €2,900 for the day, for your group |
| Including 21% Dutch VAT | €482.79 per person | €3,509 |
| The training day | Included | Included |
| A 30-minute intake call on your tools and data rules | Included | Included |
| Course materials and templates | Included | Included |
| Attendance record for your AI literacy file | Included | Included |
| Travel within the Netherlands | Not needed | Included |
| Group | 1 to 12 people | 1 to 12 people |
Prices exclude VAT. Online, live, you pay per participant. At your office, one price covers your whole group, from 1 to 12 people; for a larger group, book a second day. You reserve on this page; the payment details appear when you do.
Questions
How do I decide which tasks in my team AI should do?
List the recurring tasks, then sort each by the cost of an error, data sensitivity and how easy the result is to check. In the course you test the sort on examples from your own team.
Can managers use AI to screen CVs or rate staff?
Decisions based only on automated processing that have legal or similarly significant effects on a person fall under GDPR Article 22, and AI Act high-risk rules for employment uses apply from 2 December 2027. The course sets a team rule where AI may gather evidence and a person decides and records why.
What is AI slop and how do we keep it out of client work?
AI slop is unchecked AI output that looks finished: invented quotes, references that do not exist, filler text. The course gives your team a written quality standard and a sign-off line for every deliverable.
Does this course count towards our AI Act AI literacy duty?
It supports your Article 4 AI literacy duty. The AI Act does not require a certificate; each participant gets an attendance record for your AI literacy file.
What should we do after the course?
Start the check routine on the first 3 tasks and count the measures from the baselines. The plan sets a review date for comparing results with those baselines and moving tasks between groups.
Is this management course suitable without a technical background?
Yes. Exercises use your organisation's approved AI tool and your team's own tasks, and no coding is involved.
Last reviewed 20 October 2026.
Related courses
Leading AI Adoption: Governance, Policy and Culture
Draft your AI use policy and approved tool list on your own data, and set who decides on new AI tools and uses.
€399 per person online · €2,900 a day at your office, up to 12 people
AI for Daily Work: Prompts, Assistants and Automations
Build a tested assistant or simple automation for a recurring task, with a check step and an owner.
€399 per person online · €2,900 a day at your office, up to 15 people
AI Basics at Work
Use your organisation's AI assistant for real work, check what it gives you, and keep company data safe.
€399 per person online · €2,900 a day at your office, up to 15 people