APIs and OpenAPI in the Age of AI
summary
OpenAPI today gives paths, schemas and examples but lacks intent and workflow details, side‑effect signals, and semantic outcomes; the talk argues those gaps slow reliable AI consumption. A new year‑old spec called Orazio workflow spec already encodes multi‑call workflows so companies can execute deterministic sequences, and an overlays patch format lets platforms expose or hide subsets and annotate AI usage and model/training policies for consumers.
AI-written summary · about
chapters (10)
- 0:00APIs and AI introduction
- 1:50AI's impact on APIs
- 3:56Why API descriptions matter
- 8:09Gaps in OpenAPI for AI
- 14:30Filling gaps with standards
- 17:18Arazzo workflows for AI
- 19:05Designing AI-ready APIs
- 21:35Signaling AI in APIs
- 24:42OpenAPI's expanding ecosystem
- 26:04Future of API standards
related talks
29mOpenAPI Track Welcome and OpenAPI in the Age of AI
31mBeyond OpenAPI: API Workflows That AI Agents Can Actually Execute
40mAPI Lifecycle: Smarter Design and Governance With Open Tooling
23mDesigning APIs Both Humans and AI Love
28mOpey, the AI for API Discovery and Future of Fintech
26mAPI Governance: The Latest Trends Making Teams Love API Standardization.
43mDesigning REST APIs for the age of AI agents
27mHow AI is Transforming API Design: Case Studies & Insights