
Why AI is a management function rather than a technology tool
The responsibility, sphere of influence and indicators of the role
How it differs from the technology and data leadership roles
Where it should sit structurally, and what happens when nobody owns it
The link from business strategy to AI strategy
Use cases along the value chain
Build, buy or partner
The planning horizon, and how to present the strategy to a board
Initiatives against products
Internal against external AI
Prioritising by impact, feasibility and risk
Quick wins against long-term work, and when to shut something down
Data readiness, and what to check before anything else
What you need to understand at executive level about models and agents
In-house models against external platforms
Vendor lock-in and technology risk
Upskilling, reskilling and replacement
AI literacy for managers and employees
Resistance: fear, quiet sabotage and illusions
The team you actually need to hire
Legal, ethical and reputational risk
AI governance and the principles of accountability
Calculating the return, and what genuinely cannot be calculated
The typical failures of AI transformations
The range of use, from chatbots to full systems
The models in common use
What the research says about time actually saved
The competency model for working with this, and which systems are integrating AI
The systems offering AI in selection
Which specific functions get automated
Whether a candidate can be assessed fully automatically, and what that implies
Interview analysis tools, and the ones that raise questions
Using AI to build search queries
Sourcing plugins and automation tools
Platforms with AI built into the search
Where automated search starts touching personal data
Building a checklist with AI
Onboarding chatbots
Platforms with AI built in
What else in onboarding can be handed over, and what should not be
Systems that already integrate AI into performance management
Where this is heading with connected devices
Software that analyses performance and generates a development plan
The benefits claimed, and the questions they raise
The marketing tools applicable to an employer brand
Brand development and visual generation
Text generation
Monitoring with AI, and where generated content needs disclosure
Developing courses with AI
Building development plans
Translation, transcription and subtitling
Generating video content from text
The three categories of solution
Which data about people these systems run on
Career planning with AI
Internal mobility, and the fairness questions inside it
Developing pay policy
Designing bonus schemes
Recognition programmes
AI-assisted job grading, which is a decision with direct financial consequence
Developing culture with AI assistance
Creating stories that teach it
Selecting the artefacts
Where generated communication stops sounding like the company
Predictive analytics with AI
Predicting who will leave, which is the most sensitive use in the set
Analysing reports and drawing conclusions
Finding correlations, and doing advanced analysis without code
Creating documents and templates
Building processes
Using AI as a consultant on a problem
Financial justification of a decision with AI, and where to verify it
How decisions produce risk
What the last financial crisis demonstrated about it
What a people risk actually is
Why AI systems concentrate rather than remove this class of risk
The risk heat map, and how to build one
A worked example of risk management in this area
The fraud triangle and the safety triangle
The risk triangle applied to people and to systems acting on them
How decisions are actually made under pressure
Cognitive biases and their effect on control
Conflict of interest, and how it is normally concealed
Why automation inherits the bias of whoever specified it
Real examples of risk in this category
The 4I model
Incidents, and what they reveal about the system
Risks at individual, company and industry level
What risk management involves as a process
The ISO 31000 standard
The risk hierarchy, and internal against external risk
Finding, assessing, reducing and escalating
Management maturity as a risk factor
The dynamics inside a leadership team
The role of the chief executive
The warning signs that appear before a failure
The cultural context around any control
Building a culture where a problem is reported rather than hidden
Company values, and whether they survive contact with pressure
Trust as an operational variable
The three lines of defence model
Reducing risk at each line
What a high-reliability organisation does differently
Where the support function sits, and auditing culture
Making the invisible visible
Improving how decisions are actually made
A decision-making checklist you can apply immediately
The code of conduct and what it is meant to do
The role description as an accountability instrument
Where governance finally becomes one named person, including for a system
This course contains the use of artificial intelligence.
Most companies adopted AI before deciding who was accountable for what it does. The tools arrived department by department, and the policy is still a paragraph somebody drafted after an incident.
Why governance written first never sticks
Because it is written by people who cannot say concretely what the systems do. A policy that says "AI use must be ethical and human-supervised" survives one meeting and stops nothing, since nobody knows which decisions are already automated, which model produced them, or who would notice if the output drifted. Governance has to start from an inventory of what is actually running, and almost nobody builds one.
What this course actually covers
Five blocks, from the subject to the framework. First, ownership and strategy: who is accountable for AI in a company, how that differs from the technology roles, when AI is worth doing and when it is not, how to prioritise initiatives by impact, feasibility and risk, and one lesson devoted to risk, ethics, return and the principles of accountability. Second, twelve lessons on what AI is actually doing across a business — screening, onboarding, performance assessment, reward decisions, predictive analytics — because governance without that layer stays abstract. Third, the risk framework itself: the heat map, the fraud triangle, ISO 31000, the risk hierarchy, top-management risk, risk culture and the three lines of defence, ending with personal accountability. Fourth, bias and fairness — how discrimination operates unnoticed, which is precisely what a model reproduces at scale. Fifth, the audit: three levels, the checklist, four data collection methods, gap analysis, the responsibility matrix and the report with a return calculation.
A note on scope and on examples
Twenty-five of the thirty-seven lessons use people-related systems as their material: automated screening, attrition prediction, performance assessment. That is deliberate. Decisions about people are where AI met real consequences first, so that is where the practice and the scrutiny actually exist. The apparatus is domain-neutral — ISO 31000, the three lines of defence, a risk heat map, an audit checklist and a responsibility matrix work on any system. This course does not cover the EU AI Act or other specific legislation; it covers the governance you build regardless of jurisdiction.
Who is teaching this
I am Mike Pritula. I built the people system at Preply as it became a unicorn, and I have worked at Wargaming, iDeals and Alfa-Bank. More than 1.6 million students have enrolled in my courses across 185 countries, and over 150,000 specialists have gone through my programmes. I hold PHRi and SHRM-CP certifications and represent HRCI in more than ten countries.
What is included
Lifetime access to all 37 lessons
Active instructor support in the Q&A section
A Udemy Certificate of Completion
Working material: initiative prioritisation, the risk heat map, ISO 31000, the three lines of defence, the audit checklist and levels, the responsibility matrix, the return calculation
A realistic inventory of what AI is already doing inside an organisation
Where to start
List every AI tool anyone in your company uses, including the ones nobody approved. That list is the actual scope of your governance problem, and it is almost always longer than the policy assumes. Enrol now and start today.