CVE-2026-64636: Plesk authenticated SQLi exposes

CISO Take

This is a classic SQL injection (CWE-89) in Plesk Obsidian's web hosting control panel, letting any authenticated user — even a low-privileged reseller or customer account — read arbitrary rows from the panel's backend database. Despite the High CVSS score of 7.7 (driven by scope-changed confidentiality impact), the actual urgency signals are low: it's not in CISA KEV, CISA's own SSVC decision is TRACK (the lowest priority tier), there's no public exploit or Nuclei template, and the EPSS absolute score is just 0.24% even though that ranks in the 85th percentile of scored CVEs. Note this flags as AI-adjacent (ml_ui category) but Plesk is a general server/hosting management panel, not an AI/ML framework, model, or pipeline component — its relevance to AI teams is only indirect, via credentials or configs for AI-enabled services that happen to be hosted on Plesk-managed servers. Patch to the fixed build once available from Plesk (advisory linked below), restrict or audit low-trust panel account access in the interim, and treat this as routine infrastructure hygiene rather than an AI-specific threat.

Sources: NVD CISA KEV EPSS support.plesk.com

What is the risk?

Medium-low priority. The CVSS 7.7 rating reflects a real confidentiality impact with scope change, but exploitation requires an existing low-privileged authenticated account (AV:N/AC:L/PR:L/UI:N), and multiple independent signals — no CISA KEV listing, SSVC TRACK decision, no public exploit code, no scanner template, and a low absolute EPSS probability (0.24%) — indicate this is not under active or imminent exploitation. This is standard authenticated web-application SQLi, not a novel or AI-specific attack technique.

How does the attack unfold?

Initial Access

Attacker obtains or uses valid low-privileged Plesk panel credentials (reseller or customer account).

Exploitation

Attacker submits crafted input to a vulnerable panel parameter, triggering the SQL injection (CWE-89) against the backend panel database.

Impact

Attacker reads arbitrary rows from the panel database, exposing tenant credentials, configuration, or other sensitive hosting data beyond their authorized scope.

Initial Access

Attacker obtains or uses valid low-privileged Plesk panel credentials (reseller or customer account).

Exploitation

Attacker submits crafted input to a vulnerable panel parameter, triggering the SQL injection (CWE-89) against the backend panel database.

Impact

Attacker reads arbitrary rows from the panel database, exposing tenant credentials, configuration, or other sensitive hosting data beyond their authorized scope.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
Plesk	—	—	No patch

Do you use Plesk? You're affected.

How severe is it?

CVSS 3.1

7.7 / 10

EPSS

0.3%

chance of exploitation in 30 days

Higher than 17% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

No known exploitation

Sophistication

Trivial

What is the attack surface?

AV Network

AC Low

PR Low

UI None

S Changed

C High

I None

A None

What should I do?

1 step

Apply the vendor patch once Plesk publishes a fixed build for Obsidian (track https://support.plesk.com/hc/en-us/articles/42431868205079); until then, minimize the number of low-trust panel accounts (resellers/customers) and tighten their permissions where the panel allows. Enable and review Plesk panel audit/access logs for anomalous queries or unexpected data access patterns from non-admin accounts. If any AI-enabled service credentials or configuration are stored in or reachable via the Plesk panel database, rotate them as a precaution and verify no unauthorized reads occurred before the patch was applied.

What does CISA's SSVC say?

Decision Track

Exploitation none

Automatable No

Technical Impact partial

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

Data Extraction Data Leakage Framework

Which compliance frameworks are affected?

This CVE is relevant to:

NIST AI RMF

GOVERN 6.1 - Third-Party Risk Management

Frequently Asked Questions

What is CVE-2026-64636?

This is a classic SQL injection (CWE-89) in Plesk Obsidian's web hosting control panel, letting any authenticated user — even a low-privileged reseller or customer account — read arbitrary rows from the panel's backend database. Despite the High CVSS score of 7.7 (driven by scope-changed confidentiality impact), the actual urgency signals are low: it's not in CISA KEV, CISA's own SSVC decision is TRACK (the lowest priority tier), there's no public exploit or Nuclei template, and the EPSS absolute score is just 0.24% even though that ranks in the 85th percentile of scored CVEs. Note this flags as AI-adjacent (ml_ui category) but Plesk is a general server/hosting management panel, not an AI/ML framework, model, or pipeline component — its relevance to AI teams is only indirect, via credentials or configs for AI-enabled services that happen to be hosted on Plesk-managed servers. Patch to the fixed build once available from Plesk (advisory linked below), restrict or audit low-trust panel account access in the interim, and treat this as routine infrastructure hygiene rather than an AI-specific threat.

Is CVE-2026-64636 actively exploited?

No confirmed active exploitation of CVE-2026-64636 has been reported, but organizations should still patch proactively.

How to fix CVE-2026-64636?

Apply the vendor patch once Plesk publishes a fixed build for Obsidian (track https://support.plesk.com/hc/en-us/articles/42431868205079); until then, minimize the number of low-trust panel accounts (resellers/customers) and tighten their permissions where the panel allows. Enable and review Plesk panel audit/access logs for anomalous queries or unexpected data access patterns from non-admin accounts. If any AI-enabled service credentials or configuration are stored in or reachable via the Plesk panel database, rotate them as a precaution and verify no unauthorized reads occurred before the patch was applied.

What is the CVSS score for CVE-2026-64636?

CVE-2026-64636 has a CVSS v3.1 base score of 7.7 (HIGH). The EPSS exploitation probability is 0.25%.

What is the AI security impact?

Compliance Controls Affected

NIST AI RMF: GOVERN 6.1

What are the technical details?

Original Advisory

An SQL injection vulnerability in Plesk Obsidian up to 18.0.80 for Linux and Windows allows an authenticated user to read arbitrary data from the panel database.

Exploitation Scenario

An attacker who has obtained or purchased low-privileged Plesk panel credentials (e.g., a compromised reseller or customer account) submits crafted input to a vulnerable panel parameter, triggering the SQL injection to read arbitrary tables from the shared panel database — potentially harvesting other tenants' credentials, API tokens, or server configuration, including any metadata tied to AI workloads hosted on the same infrastructure.

Weaknesses (CWE)

CWE-89 Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection')

CWE-89 — Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection'): The product constructs all or part of an SQL command using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the intended SQL command when it is sent to a downstream component. Without sufficient removal or quoting of SQL syntax in user-controllable inputs, the generated SQL query can cause those inputs to be interpreted as SQL instead of ordinary user data.

[Architecture and Design] Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid [REF-1482]. For example, consider using persistence layers such as Hibernate or Enterprise Java Beans, which can provide significant protection against SQL injection if used properly.
[Architecture and Design] If available, use structured mechanisms that automatically enforce the separation between data and code. These mechanisms may be able to provide the relevant quoting, encoding, and validation automatically, instead of relying on the developer to provide this capability at every point where output is generated. Process SQL queries using prepared statements, parameterized queries, or stored procedures. These features should accept parameters or variables and support strong typing. Do not dynamically construct and execute query strings within these features using "exec" or similar functionality, since this may re-introduce the possibility of SQL injection. [REF-867]

Source: MITRE CWE corpus.