The crypto industry demands that AI labs open their doors: network security is at risk
Leading players in the crypto industry have joined forces in an unprecedented coalition to demand that the world's largest artificial intelligence labs provide verified open-source developers with access to their cutting-edge models. This is not merely a technical whim, but a matter of survival for the financial infrastructure that protects trillions of dollars daily.
The Bitcoin Policy Institute, together with dozens of companies—from giants like Coinbase and Block to specialized cybersecurity firms and custodial services—has published an open letter. Signatures were added by representatives of BitGo, Blockstream, Trezor, Galaxy, MARA, Foundry, and many others. They all state: network defenders who monitor code integrity remain without the most powerful tools available to attackers.
Why this is critical
Open source has long become financial infrastructure. According to the authors' estimates, the Bitcoin network alone safeguards over a trillion dollars, while the entire digital asset ecosystem—from wallets to exchanges and payment networks—manages sums in the trillions. Behind this code stand the personal savings of millions of people: pension funds, corporate capital, family wealth. A single vulnerability could destroy everything.
The problem is compounded by the fact that advanced AI models are already reshaping the economics of cyberattacks. Attackers are likely using them for continuous assaults, while small teams of maintainers are forced to rely on less powerful open-weight models. Bitcoin Core developers and other security researchers find themselves in the most vulnerable position in the entire system.
Specific demands
The coalition asks labs to create or expand trusted access programs for qualified defenders of open financial infrastructure. This involves early, controlled access to the strongest cyber models, including pre-release versions, as well as "sufficient" computing power for agents to work autonomously on long-running tasks. Without this, comprehensive security audits are simply impossible.
Special emphasis is placed on the need for secure environments to analyze closed-source code, as well as the right of small non-profits and independent maintainers to participate on equal footing with large corporations and states. A direct channel for coordinated vulnerability disclosure and communication with lab security teams is another key point.
The digital asset industry is ready to help identify trusted participants, establish standards, and coordinate remediation efforts. Cutting-edge AI could become one of the most powerful defensive technologies in history, but only with fair access to such systems.
This letter is a signal that the arms race in cyberspace has entered a new level. While AI labs view their models as commercial assets, defenders of open infrastructure see them as the only chance to hold the line. A compromise is inevitable; otherwise, the cost of a mistake will be measured not in bitcoins, but in people's lives.
My take: this step was long overdue. Without access to top-tier AI models, the open-source community is doomed to play catch-up. Labs that take the first step toward cooperation will not only strengthen the security of the entire ecosystem but also gain a reputational advantage that is hard to overstate.