The API lifecycle: No shortcuts, even with AI

The engineering fundamentals you need before scaling AI across API governance, documentation, and testing.

In SmartBear’s State of Software Quality and Testing 2026 survey, 55% of software professionals reported application quality issues in the past 12 months because AI-accelerated development outpaced testing. Only 22% of organizations have independent checks at all three points: before AI generates code, before commit, and before release.  

In this session, Mark Boyd from Future of Software Technologies (FOST) and Joe Joyce from SmartBear walk through the AI-enabled API lifecycle, built on three principles: humans steer, rules run, and AI assists. You’ll see where AI speeds things up, where deterministic tools like linting and contract testing should make the pass/fail calls, and where human accountability can’t be delegated. Done right, governance gets you to market faster, with less rework and more confidence. 

What you’ll learn: 

• How to solve the “dumbbell problem,” where the hardest work sits at both ends of the lifecycle: defining intent and governing at scale 
• How to split work between humans, rules, and AI across lifecycle stages 
• Why pass/fail decisions belong in rulesets, not agents, and how that lowers cost and risk 
• How to prepare for 10x the agent use, with drift detection and test scenarios for when agents should be stopped 
• Steps to adopt this model 

Who should attend:
API and platform engineering leaders, architects, and product owners responsible for API strategy and governance.