Crypto Developers Seek Frontier AI Access to Counter Cyber Threats

The Bitcoin Policy Institute, supported by firms including Kraken, Ledger, and BitGo, argues that current safety restrictions on frontier AI models inadvertently disadvantage the defenders of critical financial systems. While developers maintain software securing over $1 trillion in value, they are often relegated to using less capable, open-weight models. This creates a dangerous asymmetry where malicious actors—potentially backed by foreign adversaries—leverage superior AI to identify and exploit vulnerabilities at scale.

To bridge this gap, the coalition proposes standing trusted-access programs for qualified security researchers. The goal is not to remove safety guardrails for the public, but to create a vetted pathway for those responsible for maintaining foundational codebases. Recent history underscores the urgency; Bitcoin Core developers have faced high-severity bugs, such as CVE-2024-52911, which required costly, manual intervention to resolve. Proponents suggest that integrating frontier models into defensive workflows could allow for the rapid inspection of massive codebases, effectively turning AI into a potent shield against the industry's recurring security crises, which saw over $634 million stolen in a single month in 2026.