Crypto Firms Ask AI Companies for Early Access to Bitcoin Devs

A coalition of crypto companies and industry groups has asked frontier artificial intelligence (AI) labs to provide Bitcoin developers and other open-source “defenders” early access to their most capable models. The request comes in a letter published Monday by the Bitcoin Policy Institute (BPI), arguing that current access arrangements can leave critical infrastructure teams operating behind the pace of rapidly advancing AI-assisted cyber capabilities.

In the letter, signatories say many defenders—including Bitcoin Core developers—can be limited by the absence of dedicated “trusted-access programs” and by guardrails applied to publicly available frontier systems. As a result, they contend that qualified teams may be forced to rely on less capable open-weight models, even as attackers may use more powerful AI tooling to probe for weaknesses.

Key takeaways

• The Bitcoin Policy Institute letter calls for “standing trusted-access programs” so open-source financial infrastructure defenders can use top-tier frontier AI before widespread public release.
• Signatories argue that guardrails and limited access to advanced models can hinder security research and response for Bitcoin and broader crypto systems.
• The letter links the push to the rising scale of AI-enabled vulnerability discovery and threats, citing multiple reports from open-source maintainers.
• Industry data referenced in the letter points to a sharp jump in monthly crypto hacks, with April 2026 losses exceeding $634 million.
• The coalition includes major ecosystem participants such as Anchorage Digital, BitGo, Bitwise, Blockstream, Kraken, Ledger, and Trezor, among others.

Why the letter centers on “trusted access”

The BPI says the economics of security research and cyber operations are shifting as frontier AI models become more capable. According to the letter, advanced systems can search large codebases, surface potential weaknesses, and compress timelines for complex technical work—benefits that apply not only to attackers, but also to defenders responsible for maintaining open-source financial infrastructure.

Without early, dedicated access programs, the letter warns that defenders may struggle to keep pace with evolving threats. It also argues that cyber incidents exploiting open-source vulnerabilities can translate directly into real-world harm, including the risk of losing “life savings,” given how widely open-source software underpins digital finance.

To address this asymmetry, the letter asks frontier AI labs to “establish or expand standing trusted-access programs” for qualified open-source defenders. The focus is less on broad public access and more on structured access channels for teams charged with safeguarding infrastructure.

What data and security commentary are used to support the case

The letter points to recent increases in hack activity across the sector. It cites DefiLlama data showing that total monthly crypto hacks surged in April 2026, with malicious actors stealing more than $634 million from cryptocurrency platforms—described as the highest monthly total since the Bybit hack. That earlier incident, the letter notes, contributed to losses of roughly $1.4 billion in February 2025, again according to DefiLlama.

In addition to incident volume, the letter frames AI as a force multiplier for vulnerability discovery. It references concerns raised across the crypto security industry as newer AI systems make it easier to automate parts of the probing and exploit development cycle.

Earlier coverage cited within the letter highlights comments from Mitchell Amador, CEO of bug bounty platform Immunefi, who characterized the moment as a “vulnerability apocalypse” in relation to developments in AI-assisted research. The letter also mentions the emergence of newer frontier models—described in the article as Claude Opus 4.8 and ChatGPT 5.5—as part of the broader shift raising security stakes.

Who signed the request

The open letter is co-signed by a broad cross-section of the crypto industry, signaling that the concern is not confined to one segment of infrastructure. Alongside the Bitcoin Policy Institute, the signatories include organizations such as the African Bitcoin Institute, Anchorage Digital, BitGo, Bitwise, Blockstream, Bull Bitcoin, MARA, Kraken, Ledger, and Trezor, among others.

By bringing together companies spanning custody, exchanges, analytics, wallet infrastructure, and Bitcoin-focused organizations, the letter underscores the “system-wide” nature of the risk it describes: open-source code and shared software dependencies can affect multiple products, operators, and user bases at once.

Implications for Bitcoin developers and the broader security community

If frontier AI labs establish or expand trusted-access programs as requested, the most immediate practical impact would be on the speed and effectiveness of defensive work around open-source financial infrastructure. In the letter’s framing, having early access to capable models could improve how maintainers audit code, identify potential weaknesses, and respond to new exploit techniques.

The request also highlights a tension that many in security research recognize: attackers may benefit from advanced tools faster than defenders can. By arguing that public guardrails and limited availability of powerful models can block legitimate defense work, the letter effectively calls for a policy-like solution—one that treats certain defenders as authorized users of frontier capabilities.

At the same time, it remains unclear what “standing trusted-access programs” would look like in practice, including how labs would vet applicants, what models would be shared, and how output would be handled. The letter is a policy request rather than a technical specification, so builders and investors should watch for follow-up actions that clarify implementation details.

For now, the key signal is the coalition’s insistence that time-to-defense matters as AI capabilities scale—especially as hack activity remains elevated and AI-assisted vulnerability discovery accelerates. The next phase will likely involve whether frontier AI labs respond, and whether any program structures emerge that could help Bitcoin and other open-source maintainers close the gap between defensive capacity and adversarial capability.