# CVE-2026-44467: Anthropic Claude Desktop SSH host key verification bypass

Source: https://junglewise.ai/threats/cve-2026-44467-anthropic-claude-desktop-ssh-host-key-verification-bypass  
Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats  
Published: 2026-05-13. Updated: 2026-06-02  
License: CC BY 4.0 with a link back

CVE-2026-44467 · Severity: medium · CVSS 6.8 · Published 2026-05-13

Claude Desktop is an application used for AI-assisted coding and remote development. A security flaw in its SSH feature allowed attackers on the same local network to intercept and modify remote development sessions. This could lead to the theft of sensitive code or the injection of malicious commands into the developer's environment.

Technologies: [Anthropic Claude Desktop](https://junglewise.ai/threats/technologies/claude-desktop). Vendors: [Anthropic](https://junglewise.ai/threats/vendors/anthropic).

## Quick answers

**Is CVE-2026-44467 exploited in the wild?** No. Junglewise Threat Intelligence has no report of CVE-2026-44467 being exploited in the wild.

**How severe is CVE-2026-44467?** Medium severity, with a CVSS score of 6.8, published 13 May 2026.

**Which products and versions are affected?** Anthropic Claude Desktop 1.2581.0 to 1.4304.0, from 1.2581.0 before 1.4304.0.

**Is there a fix for CVE-2026-44467?** No fixed version is recorded yet. Check the vendor's advisory in the references below. If you run it, Junglewise drafts a mitigation that fits your setup and applies it once you approve, with senior security engineers behind it.

**How is CVE-2026-44467 attacked?** It is attacked from the adjacent network, in a network appliance.

**Am I affected by CVE-2026-44467?** Check whether you run Anthropic Claude Desktop at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

## Technical details

Claude Desktop's SSH remote development feature (versions 1.2581.0 to 1.4304.0) improperly validated server identities. While the application checked if a hostname existed in the user's known_hosts file, it failed to compare the server's presented host key against the stored key. An attacker positioned on the local network (e.g., via ARP spoofing or rogue Wi-Fi) could present an arbitrary SSH host key, which the application would silently accept. This allows for a Man-in-the-Middle (MitM) attack, granting the attacker the ability to intercept or modify traffic within the SSH session. The vulnerability is resolved in version 1.4304.0.

## Affected products

- Anthropic Claude Desktop from 1.2581.0 before 1.4304.0

## Timeline

- 2026-05-06: advisory: Vendor advisory published on GitHub
- 2026-05-13: disclosed: CVE published to NVD

## References

- [https://github.com/anthropics/claude-code/security/advisories/GHSA-3rwf-2g6p-c2f9](https://github.com/anthropics/claude-code/security/advisories/GHSA-3rwf-2g6p-c2f9)

## Related threats

- [Anthropic Claude Desktop arbitrary code execution via uncontrolled search path](https://junglewise.ai/threats/cve-2026-88066-claude-desktop-is-anthropic-s-desktop-application-for-interacting)
- [Anthropic Claude Code plugin hash pin bypass in git checkout](https://junglewise.ai/threats/cve-2026-86063-claude-code-plugin-hash-pin-bypass-allows-bypass-of-hash-pinning)
- [Anthropic Claude Desktop Cowork missing integrity check in VM image handling](https://junglewise.ai/threats/cve-2026-7574-anthropic-claude-desktop-cowork-missing-integrity-check-in-vm)
- [Anthropic Claude Desktop privilege escalation in CoworkVMService](https://junglewise.ai/threats/cve-2026-44470-anthropic-claude-desktop-privilege-escalation-in-coworkvmservice)
- [Anthropic Claude for Windows privilege escalation in installer](https://junglewise.ai/threats/cve-2026-22561-anthropic-claude-for-windows-privilege-escalation-in-installer)