Matthew Skelton, co-author of Team Topologies, argued at QCon London 2026 that 80% of firms see no tangible AI benefit due to organisational immaturity rather than technical shortcomings. He proposed that Team Topologies principles serve as 'infrastructure for agency' — particularly bounded agency, where AI tools operate within constrained guardrails similar to how human authority is scoped. Skelton warned against granting AI unbounded data access (referencing OWASP LLM06 Excessive Agency), drew parallels between human cognitive load and AI context windows, and introduced the Innovation and Practices Enabling Team concept for scaling internal knowledge. Real-world examples include JP Morgan reducing 60% of dependencies via opt-in adoption and EBSCO saving $9.1M annually through optimised delivery. Skelton's forthcoming book, Adapt Together, aims to operationalise value flow for the AI era.
Sort: