When AI Creates New Forms of Technical Debt — and How Teams Manage It

AI-assisted development is generating a new class of technical debt that traditional tracking, ownership, and review practices weren't designed to catch, and across engineering teams adopting these tools, common patterns are surfacing in places teams aren't used to looking. Code is being accepted faster than it's fully understood. Subtle drift appears between prompts, models, and architectural intent. Bottlenecks emerge in onboarding and incident response rather than in delivery throughput. Even when humans own every accepted suggestion, the volume and plausibility of AI proposals shifts where attention needs to be paid. This roundtable opens space for an honest peer conversation about what these new debt patterns look like in practice, how they differ from the technical debt teams have managed for decades, and where ownership and review practices may need to evolve to keep human judgment meaningfully engaged. Participants will leave with sharper language for what they're seeing in their own teams and a candid view of where current practices fall short.