Is CVE-2026-72898 (Metabase SQL Injection) being exploited in the wild?

Yes. CISA added CVE-2026-72898 to its Known Exploited Vulnerabilities catalogue on 11 August 2026, which it does only when there is reliable evidence of active exploitation. CISA has not linked it to known ransomware campaigns. Metabase contains a SQL Injection vulnerability that allows an unauthenticated remote attacker to inject arbitrary SQL into the Metabase application database, which can give them administrator access to the instance.

What makes this a confirmed in-the-wild exploitation case?

CISA’s KEV listing is the key signal here: it is used for vulnerabilities with reliable evidence of active exploitation, so this is not just a theoretical SQL injection issue. The practical question for defenders is whether Metabase is exposed anywhere in the environment and whether its application database or admin path can be reached from an untrusted network.

Because the flaw is unauthenticated and remote, the attack path is straightforward: an external actor can inject SQL into the Metabase application database and potentially reach administrator-level control of the instance. That means exposure assessment should focus on internet-facing deployments first, then on whether any reverse proxy, WAF, or network filtering actually constrains the vulnerable endpoint.

If you are triaging this as part of a broader vulnerability queue, use the Zero-Day and Exploited Vulnerabilities Guide to align KEV-driven response with time-sensitive containment, and use the CISA Known Exploited Vulnerabilities Catalog as the authoritative exploitation signal and remediation reference.

How should teams verify exposure and contain the flaw quickly?

Start by confirming every Metabase instance, including shadow deployments, test copies, and externally reachable appliances. Then verify whether the vulnerable version is present in any path that can accept HTTP requests from users, partners, or the public internet. For a SQL injection issue, configuration drift matters as much as patch state, because a forgotten instance can remain reachable long after the main fleet is remediated.

Containment should be direct: remove external reachability where possible, restrict access to trusted management networks, and apply the vendor fix or mitigation as soon as the approved advisory permits. Where patching is delayed, the immediate goal is to reduce exploitability, not to rely on detection alone. For this class of issue, a reverse proxy rule or WAF control can buy time, but it should be treated as temporary risk reduction rather than a substitute for remediation.

Use the SQL Injection Prevention Guide when you need to validate how the flaw maps to query handling, and consult the Database Security Guide for the access controls and auditing that limit blast radius if SQL execution is abused.

What should be checked for signs of prior compromise?

Because exploitation can yield administrator access to the Metabase instance, incident review should include application logs, proxy logs, database audit trails, and any signs of newly created accounts, altered permissions, or unusual query patterns. Look for configuration changes, dashboard or collection access that does not fit normal admin activity, and any evidence that the application database itself was modified in ways that support persistence.

If the instance handled secrets, embedded credentials, or integration settings, treat those as exposed until proven otherwise. A successful SQL injection against an admin-capable application can often become a pivot into connected systems, so compromise assessment should extend beyond Metabase itself. That is especially important if the instance stores database connections, API tokens, or other authentication material in reachable configuration tables.

For response discipline, the Vulnerability Management Guide is the best internal reference for exception handling, verification, and remediation tracking, while the NIST National Vulnerability Database remains useful for record reconciliation and inventory correlation across affected assets.

Risk and Threat Considerations

This vulnerability is high risk because unauthenticated SQL injection against an application that can control its own database often turns a single request into full application compromise. The main threat is not just data access, but control of the Metabase instance itself, which can expose connected data sources and administrative functions.

Failure mechanism: The attacker injects SQL through a reachable request path, abuses the application’s database trust boundary, and leverages resulting control over stored application data or admin state to expand access.

Impact: Successful exploitation can expose data, alter configuration, create or elevate administrative access, and provide a foothold for follow-on compromise of integrated systems.

Standards & Framework Alignment

This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.

CIS Controls v8, OWASP ASVS and NIST SP 800-53 Rev 5 set the governance and control requirements practitioners need to meet.

Framework	Control / Reference	Relevance
CIS Controls v8	CIS-5 — Account Management	Applies because admin exposure and account creation are central to exploitation impact.
Recommendation — Review and revoke exposed administrative access paths, then enforce least-privilege account controls.
OWASP ASVS	V8 — Authorization	SQL injection can bypass expected access boundaries and lead to admin-level actions.
	V4 — API and Web Service	The flaw is reached through web request handling and injection into application queries.
Recommendation — Validate authorization checks around all state-changing application paths and admin functions. Review request handling and server-side query construction for injection exposure.
NIST SP 800-53 Rev 5	SI-10 — Information Input Validation	Directly addresses preventing unsafe input from reaching SQL execution paths.
	AU-6 — Audit Review, Analysis, and Reporting	Compromise checks depend on log review and suspicious activity analysis.
Recommendation — Enforce server-side input validation and parameterized query handling. Review logs for abnormal queries, admin changes, and signs of exploitation.

Practitioner Guidance

What to prioritise: Treat internet exposure and active remediation as the first decision point. If an instance is externally reachable and not yet fixed, it should be handled as an urgent containment item rather than a routine patch ticket.

What to verify: Confirm whether any Metabase instance can still process requests from untrusted networks, whether the vendor fix is applied everywhere, and whether the application database contains credentials or integrations that would widen the blast radius if abused.

Practitioner takeaway: KEV-listed SQL injection issues demand response based on exploitability, not just severity labels, and the safest assumption is that every exposed Metabase instance must be validated for compromise before it is trusted again.