Bitcoin Core Developer Niklas Gögge: Relying on AI to Preemptively Find Vulnerabilities Is Not a Sustainable Security St

Bitcoin Core developer Niklas Gögge warned that AI-driven vulnerability scanning is altering the security landscape, as large language models significantly lower the cost for attackers to identify potential flaws. While over 1,000 reports have been generated for Bitcoin Core by groups like Project Loupe and the Bitcoin Red Team, no high-risk or critical vulnerabilities have been found yet. Gögge emphasized that relying on more powerful models to outpace attackers is not a sustainable strategy. Instead, he advocates for building robust testing infrastructure—including automated, fuzz, and property-based testing—to prevent entire classes of vulnerabilities. Bitcoin Core components have already undergone over 100 years of cumulative CPU fuzz testing.