Study PSM-AI Essentials AI in Scrum Work: key concepts, common traps, and exam decision cues.
The exam often asks whether AI use is helping Scrum happen better or quietly replacing the practices Scrum depends on. Stronger answers use AI to support clarity, preparation, and learning, not to bypass events, people, or evidence.
| Use of AI | Stronger or weaker? | Why |
|---|---|---|
| Drafting a facilitation agenda | stronger | supports preparation without taking over the event |
| Summarizing retrospective themes for discussion | stronger | helps inspection if the team still validates the pattern |
| Deciding Sprint priorities automatically | weaker | bypasses Product Owner and team judgment |
| Declaring work done based on generated status text | weaker | weakens transparency and quality control |
flowchart LR
A["Possible AI use"] --> B["Does it support a Scrum activity?"]
B --> C["Does it preserve human accountability?"]
C --> D["Can the team validate the output?"]
D --> E["Use carefully with review"]
If the answer becomes “no” at any step, the stronger exam instinct is usually to narrow the use, add review, or avoid that use entirely.
If AI proposes possible Retrospective themes, the Scrum Master may use that draft to speed preparation. The stronger move is still to let the team inspect and challenge the pattern, not to treat the AI draft as the conclusion.
A Scrum Master wants AI to monitor daily team messages, rewrite them into an executive update, and recommend which backlog items should be dropped from the Sprint. The stronger answer usually rejects the prioritization part first. AI can assist with summaries, but once it starts changing Sprint intent or product decisions without the right accountable people, it is no longer just support.
Which use of AI best aligns with Scrum?
A. Helping prepare a Sprint Retrospective while leaving team inspection and improvement decisions to the team
B. Reordering the Product Backlog without Product Owner review
C. Replacing the Daily Scrum with a generated summary sent to management
D. Marking items done automatically once AI predicts low risk
Best answer: A
Why: AI can support preparation and pattern recognition, but Scrum still depends on human inspection, discussion, and accountability.
Why the others are weaker: B, C, and D all replace the very mechanisms Scrum uses to stay empirical.