AI might cut false positives, but it won’t stop the slop 

The report reviews how increasingly capable AI models and open-source scanners are driving a large volume of often unvalidated or speculative bug reports that burden triage processes; GitHub is tightening its definition of a "complete" report, Cloudflare found mixed results when testing Anthropic's Mythos (which can chain exploits and generate PoC code), and the curl maintainer reported mostly false positives, concluding Mythos provided limited practical improvement.