Bug Finding Considered Harmful?

Property-based testing researcher Federico Cassano filed more than 500 bug reports in a single month, mostly against PostgreSQL, and is now questioning whether bug-finding research delivers value proportional to the burden it places on maintainers. Writing on his blog, Cassano argues that testing research’s real value proposition is building reliable software, not maximizing bug counts, and that the structural gap between outside testers and system maintainers is widening in the AI era. He notes that “fuzzer fatigue” is not new—Z3 prover drew heated debate as a fuzzing target, and SQLite maintainers have long managed heavy report volume—and worries his reports could delay PostgreSQL 19. The tension matters because AI-assisted testing is scaling report volume far faster than maintainer capacity to triage it. Watch for how major open-source projects adjust contribution norms and triage processes, and whether testing tools shift toward prioritizing and deduplicating findings rather than simply surfacing more of them.