Updated
Will AI make my Scrum team deliver faster?
Not by itself. AI can help developers write code faster, but code review, testing and release still run at human speed, so the bottleneck moves downstream. Your team delivers faster only when working software reaches real users sooner. More code written is not the same thing, and it can pile up as unfinished work.
What the Scrum Guide says
The Scrum Guide does not mention AI. It does say what makes work count, and what makes a team's work visible:
"Work cannot be considered part of an Increment unless it meets the Definition of Done."
"Artifacts that have low transparency can lead to decisions that diminish value and increase risk."
"Transparency enables inspection. Inspection without transparency is misleading and wasteful." Read the Increment section of the Interactive Scrum Guide.
Go deeper
The full argument is in my video, Why AI Won’t Speed Up Your Scrum Team (The 1999 Trap), and on its resource page. For what a team can do about it, read Code Got Cheap. Accountability Didn’t.. This page is the short answer. For the playbook behind Part 2, read What does Anthropic’s AI-native SDLC playbook mean for Scrum teams?.
What I see in practice
I call it the 1999 trap. I started writing code in 1991 and lived through two startups in the late 1990s, one IPO and one dot-bomb crash. In the video the pattern is “speculative promises first, delivery economics later.” Today’s version: “AI tools don’t remove delivery bottlenecks. They move them.”
The extra code has to land somewhere. As I wrote on that page, it lands in “peer review, the test suite and Sprint Backlog refinement, and those still run at human speed.” In Synthetic Speed I put it plainly: “Generating code faster when code reviews take two weeks does not deliver value. It just creates traffic jams and burnout.”
Part 2 of the video, Code Got Cheap. Accountability Didn’t., ends with this: “AI didn’t make the old process obsolete. It made it impossible to hide.” If your work already waited in handoffs and approvals, AI shows you exactly where.
Common mistakes
- Measuring how fast the team generates code, drafts or tickets, instead of how fast working software reaches users.
- Speeding up the top of the funnel and leaving review, testing and release exactly as they were.
- Letting work into the Sprint that nobody can inspect. In the video I say that if the team “cannot inspect the code or the requirement with their own eyes, you cannot adapt.”
- AI theater: in the video, “generating 200 backlog tickets nobody asked for.”
- Treating the tool as the one who is accountable. “Accountability remains 100% human.”
What to do this week
- Look at your last Sprint’s board. Find where work waits (review, testing, release) and write down how many days it waited.
- Ask the team in your next Retrospective how AI changed its work last Sprint, and what worked and what didn’t. AI for Scrum Teams has two questions you can borrow.
- Pick one downstream bottleneck and clear it before the team generates more. In Synthetic Speed my advice is: “Inspect your end-to-end delivery funnel, find where work accumulates, and focus your team on clearing downstream bottlenecks before generating more top-of-funnel noise.”
FAQ
Does the Scrum Guide say anything about AI?
No. The Guide does not mention AI. It does say that "Work cannot be considered part of an Increment unless it meets the Definition of Done." My take: that line is the test for AI-assisted work, the same as for any other work. See How should the Definition of Done change when developers use AI?
Where does the extra work pile up?
In the steps that still run at human speed. In the video I name peer review, the automated test suite and Sprint Backlog refinement. In Synthetic Speed the scenario was a team that generated 40 pull requests in a week while the release was completely stuck.
So should my team stay away from AI?
No. The point is where the speed shows up. In AI for Scrum Teams I cover how Product Owners, Developers and Scrum Masters can use it. My take: use it, and inspect the result in your Retrospective like any other change to how you work.