The conversation around AI in Scrum has shifted from a fringe topic to an urgent necessity, as AI is already changing how teams collaborate.
High adoption of AI tools contrasts with low developer trust, creating an invisible influence on team decisions and a gap with traditional Scrum practices.
The DORA State of DevOps report reveals that AI magnifies existing team dynamics, amplifying both high-performing habits and underlying dysfunctions.
The Scrum Master's focus shifts from solely facilitating human interaction to orchestrating human-AI collaboration and removing algorithmic impediments.
Orchestration is broken down into four practical responsibilities: designing task boundaries, creating working agreements, inspecting AI's impact, and ensuring psychological safety.
Use patterns like "AI drafts, human decides" to intentionally define which tasks are delegated to AI and which require human ownership and judgment.
Establish explicit team agreements, such as treating AI output as a first draft and maintaining human accountability, to guide AI usage.
Integrate specific questions into the Daily Scrum, Sprint Review, and Retrospective to inspect and adapt how the team collaborates with AI.
Address key risks like comparison pressure, junior learning loss, bias amplification, and surveillance creep to maintain team health and ethical standards.
A team experienced increased velocity but also rising instability and technical debt because their standard retrospectives failed to uncover AI's negative side effects.
The team implemented small changes like adding AI-specific retrospective questions and updating their Definition of Done to regain quality and shared understanding.
A simple model based on visibility, boundaries, and learning loops helps teams apply Scrum's core pillars to human-AI collaboration.
Start orchestrating AI in your next sprint by adding a specific retrospective question, creating one working agreement, and making AI's impact visible.
Jobs that call for the skills explored in this talk.