As reported by BleepingComputer, a newly disclosed vulnerability dubbed "*sshole" exposes a subtle but dangerous interaction between OpenSSH's authentication logic and the musl C library's crypt() implementation. The result: attackers can authenticate to accounts explicitly configured to reject password logins — including default accounts in Alpine Linux's official AWS and Azure cloud images that carry passwordless doas (a sudo alternative) privileges.

Security Impact: As reported by BleepingComputer, a newly disclosed vulnerability dubbed "*sshole" exposes a subtle but dangerous interaction between OpenSSH's authentication logic and the musl C library's crypt() implementation.

Why This Matters

This is not a typical authentication bypass. It is a semantic collision — two independently reasonable design decisions that, when combined under specific conditions, produce an exploitable security hole. OpenSSH's 2016 patch for CVE-2016-6210 introduced dummy hashing for accounts with unusable password hashes to prevent timing-based username enumeration. musl's crypt() returns * when it encounters a password exceeding its 256-byte input limit. When an account's stored hash is also * (the conventional way to disable password login), the comparison succeeds trivially.

The vulnerability has existed since 2016 and remained undetected for a decade — a stark reminder that complexity in authentication chains creates blind spots that even mature projects miss.

Affected Products and Conditions

AttributeDetail
Vulnerability Name*sshole
Root Cause CVECVE-2016-6210 (OpenSSH timing patch)
Severity (Shield53 Assessment)High — authentication bypass with potential root escalation
Affected C Librarymusl libc (NOT glibc)
Affected DistributionsAlpine Linux, OpenWrt, Buildroot, Yocto (and any musl-based distro)
Not AffectedDebian, Ubuntu, Fedora, RHEL (glibc-based); any system using PAM authentication
Exploitation ConditionsPAM disabled + SSH password auth enabled + ≥1 account with * hash + ≥1 account with valid hash
Cloud Image ExposureAlpine Linux official AWS and Azure images ship vulnerable 'alpine' account with passwordless doas
Active ExploitationNot yet observed (disclosed November 5, 2026)
Patch AvailabilityPending — check vendor advisories

Who Is Most at Risk

The exposure profile is narrower than it first appears but includes some high-value targets:

Who Is Most at Risk
Cloud-deployed Alpine Linux instances — particularly the default AWS and Azure images where the alpine user has passwordless doas access. An attacker who exploits *sshole on this account can pivot directly to root.
Embedded and IoT systems running OpenWrt or Buildroot with SSH exposed to management networks. These devices frequently lack monitoring and may never receive patches.
CI/CD build containers using Alpine as a base image — if SSH is exposed and accounts are locked with *, the bypass applies.
Kubernetes worker nodes or edge appliances running musl-based minimal OS images with SSH management enabled.

The good news: most enterprise Linux deployments use glibc (Debian, Ubuntu, RHEL, Fedora) and are inherently immune. The risk concentrates in lightweight, musl-based environments — which are disproportionately used in cloud-native, edge, and embedded scenarios where patching cycles are slower and exposure can be permanent.

Broader Implications

This vulnerability highlights an underappreciated class of risk: cross-component semantic flaws. Neither OpenSSH nor musl is individually buggy. The OpenSSH team's 2016 mitigation was sound for glibc systems. The musl team's crypt() return value is a reasonable design. But when combined under the specific configuration conditions present in Alpine's default setup, they create an exploitable path.

It is also notable that the researcher, Ozgun Kultekin, discovered this through AI-assisted fuzzing of authentication configurations — running ~17,600 test cases across malformed hashes and oversized passwords. This suggests that AI-driven security testing can surface interaction-level bugs that traditional vulnerability scanning and manual code review have consistently missed for a decade.

Shield53 Recommendations

Immediate Actions

  • Inventory and identify exposed assets: Search your environment for musl-based Linux hosts with SSH exposed. Prioritize Alpine Linux instances deployed from AWS or Azure Marketplace images. Run ldd $(which sshd) | grep musl or check /lib/ld-musl-*.so.1 to confirm the C library.
  • Audit the 'alpine' account: On Alpine cloud images, immediately check whether the alpine user exists and has * in /etc/shadow with passwordless doas. If so, treat this as a live root-exposure risk.
  • Disable SSH password authentication: Set PasswordAuthentication no in sshd_config and rely exclusively on key-based authentication. This eliminates the attack path entirely.
  • Enable PAM authentication: If password auth must remain enabled, configure OpenSSH to use PAM (UsePAM yes). PAM-based authentication is not affected by this flaw.
  • Replace * hashes with !: In /etc/shadow, change disabled-account hashes from * to ! as an interim mitigation. Verify this does not break any dependent services.
  • Restrict SSH access: Use security groups, firewall rules, or AllowUsers/AllowGroups directives to limit SSH exposure to trusted management IPs only.
  • Monitor for exploitation: Watch SSH logs for authentication events involving oversized passwords or successful logins to accounts that should be locked. Any successful auth on a *-hashed account is a strong indicator of compromise.

Longer-Term Actions

  • Track for patches: Monitor OpenSSH and Alpine Linux security advisories for upstream fixes. Given the root-escalation potential on default cloud images, expect a rapid patch cycle.
  • Reassess musl-based deployments: For organizations running Alpine in production, evaluate whether the lightweight footprint justifies the added risk surface from musl-specific interaction bugs. Consider glibc-based minimal images (e.g., Debian slim) where operational constraints allow.
  • Adopt AI-assisted configuration fuzzing: Kultekin's methodology — systematic testing of authentication interactions across malformed inputs — is reproducible. Security teams should consider similar approaches for critical authentication components.