ME / RESEARCH TOPIC

Software Testing

2 public research records in this topic.

Research note in development

A Copilot for the Test Bench: Where Could AI Help Software Testing?

Exploring careful uses of AI for test ideas, maintenance, and quality evidence

Software testing involves imagination as well as repetition. This developing research note asks where AI might help propose tests, maintain them, and explain quality evidence without becoming its own judge. The design remains provisional while its claims are compared with public research and reproducible experiments.

Research note in development

Testing the Tester: What Changes When AI Designs Software Tests?

A technical inquiry into test ideas, oracles, architecture checks, and maintenance

When AI helps design tests, it can also repeat mistaken assumptions or reward easy measurements. This developing research note examines useful roles, failure modes, and controls for AI-enabled testing. Claims about implemented systems and projected benefits remain unverified until reproducible evidence supports them.