Troy McAlpin: Three product team challenges that come with AI tooling

Published on January 8, 2026, Troy McAlpin’s essay addresses what he identifies as three structural challenges that AI tooling introduces to product teams — not the benefits, which tend to be well documented, but the organizational friction that arrives alongside them.

The first challenge is accountability for AI-generated work. When AI produces a design draft, a code module, or a user story, ownership becomes ambiguous. Teams that haven’t established review checkpoints before AI output reaches stakeholders often find themselves debugging work that no one felt responsible for catching. McAlpin’s recommendation is direct: apply identical quality standards to AI-assisted outputs as to human-created work, and assign clear owners before output moves downstream.

The second challenge is boundary blur. AI tools allow PMs to create mockups and engineers to contribute to documentation; the capabilities that define role scope are becoming accessible across functions. McAlpin observes that this can result in team members working outside their area of expertise without the validation that specialist review would normally provide. His guidance is practical: explicit acceptance criteria, domain expert reviewers designated for each output type, and centralized context management so that institutional knowledge doesn’t fragment when roles overlap.

The third challenge is that AI adoption disrupts the rhythm teams use for estimation and coordination. Frequent reorganizations — and the capacity uncertainty that comes with AI-augmented velocity — reduce the reliability of historical velocity as a planning baseline. McAlpin cites a concrete result from his own company, Atono, where tracking cycle time continuously rather than against historical sprint velocity allowed the team to double engineering throughput, from 10.5 to 18.4 story points per week, by identifying bottlenecks rather than attributing slow periods to overall team pace.

The article closes with a point about knowledge documentation during transitions. When AI-augmented workflows accelerate staffing changes — because smaller teams can carry more — the institutional knowledge that exists only in the heads of departing team members becomes a real risk. McAlpin recommends treating knowledge documentation as a deliverable, not an afterthought.

This piece is useful for product leads and engineering managers who are mid-implementation on AI tooling and finding that productivity gains are arriving alongside coordination costs they hadn’t anticipated.