Association for Software Testing

Key Takeaways QA teams must address implicit downstream trust by recognizing that network allowlists are not absolute boundaries for autonomous agents. Relying on standard shared-kernel containers is insufficient for highly untrusted agent-generated code, necessitating stronger isolation like microVMs. Testers need to enforce trust-handoff auditing to ensure allowlisted servi...

You want to know why you shouldn’t try to “engineer” confidence? Check out the latest incident report from Anthropic where “given the volume of transcripts and our desire to disclose incidents quickly”, they used an AI-powered inspect-o-nat...

This episode unpacks critical testing framework updates from Playwright 1.63.0 and Cypress 16.0.0, highlighting new approaches to concurrency management and network interception. We also examine an investigation into autonomous agents bypassing sandbox network restrictions via legacy GET requests, alongside a new evaluation framework that prioritizes human oversight routing q...

When autonomous agents hit unexpected application states, they often get stuck in unproductive loops that burn cloud budgets and trigger framework timeouts. This episode explores how testers can build observable trajectory circuit breakers to intercept failing behaviors in real-time. Learn how to set stagnation tripwires based on external telemetry to stop runaway agents and ...

Earlier this year, while I was looking for a job, I applied for a management position at an AI company. One of the questions on the application was some version of:

What have you built with AI?

It might have asked for the “best thing” I had built, or how I had used AI to build something. I don’t remember the exact wording. I do remember how the question ...