Coinbase, Block, BitGo, and several other industry leaders have issued a formal letter to frontier AI developers, urging them to grant security researchers access to advanced models without restrictive safety guardrails. The coalition argues that current AI protocols—originally designed to prevent the creation of malicious code—are ironically blocking white-hat hackers and security analysts from stress-testing systems against the very threats these models aim to prevent.
This move highlights a growing 'security asymmetry' in the digital landscape. While regulated AI labs impose strict filters on their tools, adversaries are increasingly utilizing open-source or jailbroken frontier models to craft sophisticated exploits and identify smart contract vulnerabilities. The crypto industry warns that keeping defenders 'handcuffed' by safety filters creates a lopsided battlefield where billions in capital remain at risk from AI-empowered attackers.
From a regulatory and market perspective, this development signals a shift toward treating AI as a critical infrastructure component for financial stability. Investors should monitor whether major AI labs like OpenAI or Anthropic introduce specialized 'security tiers' for verified researchers. While the request underscores current vulnerabilities, a successful outcome would significantly bolster the long-term resilience of blockchain networks and lower the frequency of high-profile hacks.