Image Credit: Shutterstock
More than 40 crypto, cybersecurity and open-source organizations are asking leading AI labs to give vetted defenders of Bitcoin and other open-source financial infrastructure early access to frontier cybersecurity models.
The Bitcoin Policy Institute published the “Defenders Need the Frontier” open letter on August 10. Signatories include Coinbase, Blockstream, Strategy, MARA, Galaxy, BitGo, Trezor and other digital asset and developer organizations. The coalition is seeking controlled access rather than unrestricted public use of advanced cyber capabilities.
Coalition Seeks Pre-Release Models, Compute and Secure Testing
The letter asks AI companies to establish or expand standing trusted-access programs for qualified defenders of open-source financial infrastructure.
Its proposal includes pre-release access to the strongest cyber-capable models, enough computing capacity for comprehensive audits, secure environments for reviewing private or embargoed code and direct channels with AI-lab security teams for coordinated vulnerability disclosure.
The coalition also wants independent researchers, nonprofits and smaller maintenance teams to qualify rather than limiting advanced access to large companies and governments.
Bitcoin Developers Say AI Safeguards Can Block Security Research
The letter argues that Bitcoin Core developers and other open-source defenders can remain outside specialist cyber programs or encounter safeguards that block legitimate work on publicly available frontier models.
That can push researchers toward less capable open-weight alternatives even when they are examining software responsible for wallets, node infrastructure and other financial systems, according to the coalition.
AnchorWatch CEO Rob Hamilton recently said OpenAI blocked his cybersecurity work after he had completed identity verification and program onboarding. Hamilton said the restriction interrupted research connected to the Bitcoin Red Team.
OpenAI and Anthropic Already Run Controlled Cyber-Access Programs
OpenAI expanded Daybreak on August 10 with two controlled-access tiers. Daybreak Blue gives approved defenders frontier models for defensive security workflows, while Daybreak Red provides GPT-5.6-Cyber and other capabilities for advanced vulnerability research, exploit validation and security testing.
Anthropic’s Project Glasswing gives selected organizations access to Claude Mythos Preview for defensive cybersecurity work. Anthropic has committed up to $100 million in model credits and $4 million in direct donations to open-source security organizations.
Those programs overlap with parts of the coalition’s proposal, but neither company has publicly adopted its requested framework specifically for Bitcoin and open-source financial infrastructure defenders.
The next step is whether frontier AI labs broaden access to include more qualified Bitcoin maintainers under their existing controlled cybersecurity programs.
