AI-Powered Scan Uncovers Bitcoin Core Vulnerability That Let Hackers Crash Nodes on Windows and macOS
Key Takeaways
- •An AI-driven scan using a large language model identified three Bitcoin Core vulnerabilities, with human contributors verifying the findings before they were submitted to the project's security team.
- •One flaw allowed a malicious peer to remotely crash Bitcoin nodes running on Windows, macOS, and BSD systems through memory exhaustion, while Linux systems were not affected.
- •The other two disclosed vulnerabilities involved memory growth and potential failures during blockchain reorganizations.
- •Fixes were distributed across all supported Bitcoin Core versions on October 27 and 28, 2026, prior to the public disclosure on October 28.
- •These are the first disclosures under Bitcoin Core's revised security policy, and three more medium- and high-severity flaws already fixed in version 30.0 are expected to be published in November 2026.

Bitcoin Core developers have disclosed three software vulnerabilities identified through an artificial intelligence-powered scan, including a flaw that allowed attackers to remotely crash Bitcoin nodes running on Windows, macOS, and BSD systems. Bitcoin Core is the most widely used software for operating Bitcoin full nodes — computers that independently verify transactions and blocks against the network's rules, letting operators confirm the state of the ledger without relying on third parties.
According to the disclosure, a malicious peer could overwhelm an affected node with data, triggering memory exhaustion and a crash — a pattern typical of denial-of-service flaws, in which the goal is to exhaust a system's resources until the software fails. Linux systems were not affected. Developers released fixes across all supported Bitcoin Core versions on October 27 and 28, 2026, before publicly disclosing the flaw on October 28, 2026.
The findings mark the first disclosures issued under Bitcoin Core's revised security policy, which calls for faster publication of vulnerabilities that AI tools can readily discover, rather than keeping them confidential for months while users upgrade. Related coverage: REALITY CHECK | Bitcoin Core Changes Vulnerability Disclosure Policy as AI Threats Grow.
The automated scan employed a large language model to examine Bitcoin Core's codebase, with human contributors verifying the findings before they were submitted to the project's security team. Alongside the remote crash flaw, two other vulnerabilities were disclosed: one involving memory growth and another concerning potential failures during blockchain reorganizations, the process by which the network briefly replaces recently added blocks with a competing version of the chain and node software must handle the transition correctly.
The episode underscores how AI is reshaping security across the cryptocurrency ecosystem, making it easier to identify software weaknesses while simultaneously reducing the window developers have to patch them before malicious actors independently discover and exploit the same flaws. Related coverage: REPORT | AI is 2x More Effective at Exploitation Than Detection in Crypto, Says Binance Research.
Bitcoin Core developers also said that three additional medium- and high-severity vulnerabilities, fixed in version 30.0, are expected to be disclosed in November 2026 — meaning operators already running that release hold the fixes ahead of publication, and the disclosures will offer a further look at how the revised policy performs in practice.
Source: BitcoinKE