The CVE identifier carrying a 2026 date is anomalous but not disqualifying—similar artifacts have appeared in embargoed disclosures and coordinated publications. What matters far more than the metadata is the substance: an SSRF vulnerability in Azure AD/Entra ID with a 9.9 CVSS score, specifically gated behind authenticated access. That constraint is the most important detail in the entire record, and it's being overlooked in most prioritization frameworks.
The "authorized attacker" qualifier fundamentally changes the threat model. This isn't an unauthenticated remote exploit—it's a privilege-escalation path where the adversary already holds valid Azure AD credentials, typically a service principal, managed identity, or user account. In practice, this means the prerequisite access for exploitation already exists in your tenant if you run any automation, CI/CD pipelines, Azure Functions, or custom integrations that authenticate to Entra ID. The vulnerability doesn't need to be discovered in the wild; it needs to be chained with existing credentials that are already present in your environment.
More critically, the CVSS 9.9 score measures the wrong blast radius. Azure AD doesn't just authenticate users—it issues tokens granting access to Azure Resource Manager, Microsoft 365, Intune, third-party SaaS via SAML/OAuth, and customer-built applications. An SSRF here isn't a 9.9 impact on a single workload; it's potential access to the credential-issuance pipeline itself, which means access to everything that pipeline vouches for. A CVSS 8.0 in the identity layer that exposes service principal tokens to cross-tenant exfiltration is categorically more dangerous than a 9.9 in an isolated application.
Defensive action: audit your tenant for legacy Entra Connect sync connectors, unconstrained OAuth application permissions, application proxies, and any automation identities with broad permissions. The remediation scope for this class of vulnerability extends beyond patching—plan for rotating service principals that may have hit the vulnerable endpoint, auditing automation scripts for token handling, and potentially invalidating issued tokens across affected paths. This is a remediation timeline measured in months, not days.
Finally, treat the severity rating as a floor, not a ceiling. The CVSS framework structurally fails to account for the interconnectedness multiplier that makes identity provider vulnerabilities uniquely destructive regardless of the numerical score.