Presentation: Platform Engineering in the Age of AI

Details

InfoQ's Platform Engineering in the Age of AI roundtable brings together practitioners from DKB (German bank, 5M customers), Harness (Field CTO Stephen Cihak), Sevdesk/Shine, and independent consultants. The 61-minute session covers how platform teams are responding to the pressure to manage, standardize, and govern AI tooling across engineering organizations. The core operational tension is standardization versus developer autonomy. Platform teams are being asked to provide AI tools as managed services — with cost controls, security policies, and compliance hooks baked in — while developers want flexibility to adopt whatever tools improve their velocity. The panel draws on experience from organizations that have had to make concrete choices about what belongs in the platform layer versus what remains developer-controlled. Security governance emerges as the primary driver for centralizing AI tooling. Unmanaged AI tool adoption creates data exfiltration risk: developers using external LLMs to process internal or customer code can expose sensitive information to third-party services with no audit trail. Heavily regulated organizations like DKB face additional compliance mandates that make ad-hoc AI tool adoption untenable. Cost governance is the secondary driver. Platform teams report that ungoverned AI tool sprawl creates unpredictable infrastructure spend, particularly as usage-based pricing on LLM APIs scales non-linearly with team adoption. The session is a practical reference for platform engineers starting to formalize AI tooling governance frameworks, though it contains no novel research or product announcements.