Ai For Program Management

What you'll learn

---

Tell the difference between AI as an enabler of your programme and AI as your programme's deliverable - and know which one you are in
Ground a model in your actual plan documents so it stops inventing plausible milestones, clauses and dates
Screen AI proposals against the seven problem patterns and kill the ones with no measurable baseline before they become permanent pilots
Build a benefits baseline where every efficiency metric is paired with a quality guard, so you never repeat the Klarna failure
Write a decision-rights matrix that says which decisions a machine may make - decided in advance, not argued about in the moment
Extract dependencies from unstructured plan documents with a quoted source sentence for every finding
Run AI-assisted pre-mortems, scenario design and estimate challenges that produce defensible ranges instead of confident numbers
Communicate one set of facts to a board, a works council and a regulator without the numbers quietly disappearing
Measure benefit-at-risk lead time - a programme measurement that was impractical before
Evidence EU AI Act Article 4 literacy for employees and contractors, and route Article 50 disclosure decisions correctly
Requirements

---

Experience running programmes or large projects - this is written for practitioners, not beginners
No technical or data-science background required
Access to any general-purpose AI assistant is useful for the assignments, but not required to follow the course
Description

---

This course contains the use of artificial intelligence.
There are two completely different jobs hiding behind the phrase "AI and programme management", and most of the confusion in this field comes from people not noticing which one they are in. You can use AI as an enabler of a programme that has nothing to do with AI. Or the AI can be the thing your programme delivers. This course is unambiguously the first - and it says so on screen in lecture one, because that single distinction is what separates programme managers who get real value out of this technology from the ones who spend a year running pilots and have nothing to show a board.
MIT's Project NANDA found that roughly 95% of enterprise generative-AI pilots produce no measurable business return, and the researchers attributed the divide to approach rather than model quality. That makes it your problem - which is good news, because it means it is fixable from where you sit.
You will follow one programme the whole way through. Meridian Grid is a European electricity distribution network operator: 11,400 people, four countries, and a €480M six-component transformation called Grid Intelligence. Every concept gets applied to it. By the final lecture you will know that programme well enough to have opinions about it - including about the decisions its director got wrong.
What makes this course different
Benefits realisation and change management are covered properly. Both are absent from every comparable course on this platform, and both are where programme management is actually distinct from project management.
Twelve documented prompts, not a thousand loose ones. Each library entry carries its instruction, required input, known weakness, verification step and owner. The known-weakness field is what separates a library from a pack.
Six downloadable artefacts you can defend in a governance meeting - a benefits baseline sheet, two grounded prompt libraries, a decision-rights matrix, a KPI dashboard sheet and an AI register that does double duty as your Article 4 literacy evidence.
Current on the law as it actually stands. PMI's June 2026 AI standard, the Digital Omnibus deferrals, what moved to December 2027 and - importantly - what did not move at all.
Real cases, not vendor anecdotes. Klarna's reversal. Moffatt v. Air Canada. The MIT pilot study. Each read as a programme-management failure rather than a technology story.
Who this is for. Programme managers, senior project managers moving into programme roles, PMO leads, and portfolio managers who need to answer a sponsoring board's questions about AI with something more durable than enthusiasm. No technical background is assumed - lecture two gives you the entire vocabulary you need in four words.
What you will not find here. No prompt-pack padding. No claim that any framework is a compliance shield. No worked example where everything goes right - the closing lecture leaves Meridian with a component running late and a conformity workstream that is harder than anyone estimated, because that is how real programmes go.
Eight sections, 44 lectures, eight quizzes, seven assignments with fully worked model solutions, and a final practice exam. You will leave with artefacts, not enthusiasm.
Programme managers who need to answer a sponsoring board's questions about AI with evidence rather than enthusiasm,Senior project managers moving into programme roles,PMO leads responsible for standards, governance and reporting across a programme,Portfolio managers deciding which AI initiatives get funded and which get stopped,Anyone preparing to align a programme with PMI's AI standard, the NIST AI RMF or ISO/IEC 42001
Homepage

---