IBM watsonx.data CVE-2025-36327

Exploitation requires a valid authenticated session with at least low-privileged access to the IBM watsonx.data Intelligence application - unauthenticated access is not sufficient (CVSS PR:L). Affected versions are specifically 5.2.0, 5.2.1, 5.2.2, and 5.3.0; unaffected versions are not exploitable via this vector. The attacker must be able to send HTTP or API requests directly to the application (AV:N), bypassing client-side enforcement by crafting or intercepting requests rather than using the application UI as intended. No special non-default configuration is required - the flaw is inherent to how the application enforces access controls in these versions. Exploitation is bounded by the absence of confidentiality and availability impact (C:N, A:N); only integrity actions within the application's own scope are reachable (S:U). The requirement for a valid account limits risk from external, unauthenticated threat actors.

The CVSS 6.5 base score reflects a medium-high severity driven by network accessibility (AV:N), low attack complexity (AC:L), and high integrity impact (I:H), partially offset by the requirement for low-privileged authentication (PR:L) and no confidentiality or availability impact. In real-world terms, this is not a zero-authentication flaw - an attacker must first obtain or possess a valid account. However, the low complexity (AC:L) and absence of user interaction (UI:N) mean that once authenticated, exploitation is reliable and repeatable with standard tooling. No EPSS score was provided in the available data, so exploitation probability cannot be quantified numerically. The vulnerability is not in the CISA KEV catalog and no public proof-of-concept has been identified, indicating no confirmed widespread or active exploitation at time of analysis. The primary risk is in environments where multi-tenant deployments or strict role-based access control are relied upon for data governance - a successful bypass could allow a restricted user to perform privileged data modifications or administrative actions. Integrity-only impact (C:N, A:N) limits the blast radius compared to a full compromise scenario.

An authenticated user with a low-privileged account logs into IBM watsonx.data Intelligence and routes their browser traffic through an intercepting proxy. They observe that certain administrative or data-modification API endpoints are only hidden via client-side UI controls, not protected by server-side authorization checks. By crafting a direct API request to one of these endpoints, the user successfully executes a privileged action - such as modifying a protected data asset or altering an access policy - that their role should not permit. No public proof-of-concept has been identified at time of analysis.