Open source project curl is sick of users submitting "AI slop" vulnerabilities
open-sourcesecurityai-generated-contentvulnerability-reporting
Abstraction: curl maintainer banning AI-generated fake security reports
Key points:
- Daniel Stenberg, curl's lead maintainer, declared the project is being effectively DDoSed by AI-generated vulnerability reports via HackerOne
- "We still have not seen a single valid security report done with AI help," Stenberg wrote
- New policy: reporters suspected of using AI must verify; those deemed to submit "AI slop" will be banned
- Example report claimed a "novel exploit leveraging stream dependency cycles in HTTP/3" but cited nonexistent functions and a patch that didn't apply to current code
- The AI-generated submission answered questions not asked, gave basic git instructions unprompted, and recommended hardening for tools other than curl
- curl turned 25 in 2023 and is a critical open-source data-transfer library used globally
Connections: Curl · Hackerone · Daniel Stenberg · AI Generated Content · Vulnerability Reporting