Daniel Stenberg routes curl vulnerability reports to GitHub queue

Software maintainers, labs, and major tech platforms are struggling to verify the massive volume of vulnerabilities discovered by AI tools. Projects like curl and Anthropic's scanner have shifted unvalidated or public-queue burdens onto maintainers and the public because human verification capacity cannot keep up. Meanwhile, internal enterprise telemetry shows that while AI coding adoption increases code generation output, it fails to increase finished work and creates significant code review bottlenecks. Open article →