Which versions of Snowflake Spark Connector are affected by CVE-2026-15183?

Organizations using the Snowflake Spark Connector (spark-snowflake) face critical exposure to SQL injection flaws (CVSS 9.2) enabling attackers to steal OAuth credentials and execute arbitrary SQL…. A patch is available.

How to fix or mitigate CVE-2026-15183?

Within 24 hours, identify all Spark clusters running spark-snowflake versions prior to 3.2.1 and assess their network exposure to untrusted data sources. Within 7 days, implement network segmentation to isolate vulnerable connectors, rotate all Snowflake service credentials with connector access, and enable comprehensive audit logging for Snowflake queries originating from Spark. Within 30 days, establish continuous monitoring of Snowflake security advisories for patch releases and prepare an expedited deployment process to apply updates immediately upon vendor release.

Snowflake Spark Connector CVE-2026-15183

Q: What is the CVSS score of CVE-2026-15183?

CVE-2026-15183 has a CVSS 4.0 base score of 9.2 (Critical). CVSS vector: CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X. EPSS exploitation probability: 0.2%.

EUVD-2026-43644 CRITICAL

SQL Injection (CWE-89)

2026-07-14 SNOWFLAKE

GHSA-33r9-fx8q-w69h

Authentication Bypass Privilege Escalation SQLi Snowflake Spark Connector Snowflake

9.2

CVSS 4.0 · Vendor: SNOWFLAKE

Severity by source

Vendor (SNOWFLAKE) PRIMARY

9.2 CRITICAL

CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

vuln.today AI

7.5 HIGH

Shared-session access to submit SET/staging options implies PR:L, and the required preconditions plus non-trivial injection path justify AC:H; full C/I/A impact via admin-role SQL execution.

3.1 AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H

4.0 AV:N/AC:H/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N

Primary rating from Vendor (SNOWFLAKE).

CVSS VectorVendor: SNOWFLAKE

Attack Vector

Network

Attack Complexity

High

Privileges Required

None

User Interaction

None

Scope

Lifecycle Timeline

Patch available

Jul 14, 2026 - 11:17 EUVD

Analysis Generated

Jul 14, 2026 - 09:31 vuln.today

CVE Published

Jul 14, 2026 - 08:42 cve.org

CRITICAL 9.2

DescriptionCVE.org

Multiple input validation vulnerabilities in the Snowflake Spark Connector (spark-snowflake) versions prior to 3.2.1 can allow attackers to exfiltrate OAuth client credentials, execute arbitrary SQL with the connector's Snowflake role, or redirect COPY operations to attacker-controlled storage. An attacker could exploit these vulnerabilities by supplying a crafted OAuth token request URL, placing malicious files in an ingestion pipeline, injecting SQL via staging options in a shared Spark environment , or issuing runtime SET commands in a shared Spark-SQL session to inject arbitrary SQL into the SnowflakeFallbackCatalog's option map, which executes under the cluster admin's JDBC credentials. Successful exploitation may result in credential theft, unauthorized access to Snowflake account data, or privilege escalation within connected infrastructure.

AnalysisAI

SQL injection and input-validation weaknesses in the Snowflake Spark Connector (spark-snowflake) before version 3.2.1 let attackers exfiltrate OAuth client credentials, run arbitrary SQL under the connector's Snowflake role, and redirect COPY data-loading operations to attacker-controlled storage. Reported by Snowflake, the flaws are most dangerous in shared/multi-tenant Spark clusters where injected SQL executes with the cluster admin's JDBC credentials, enabling credential theft and privilege escalation. Rated CVSS 4.0 9.2 (Critical); no public exploit identified at time of analysis.

Technical ContextAI

The Snowflake Spark Connector (spark-snowflake) is a JVM data-integration library that bridges Apache Spark and the Snowflake data cloud, using JDBC plus staged COPY INTO operations to move data in bulk and OAuth for authentication. The root cause is CWE-89 (Improper Neutralization of Special Elements used in an SQL Command), manifesting through several unvalidated inputs: OAuth token-request URLs, staging/COPY options, and connector option maps that flow into the SnowflakeFallbackCatalog. Runtime Spark-SQL SET commands and staging options are incorporated into that option map without sanitization, so attacker-controlled strings are concatenated into SQL executed over the cluster's shared JDBC session, and COPY targets can be re-pointed to arbitrary storage locations.

RemediationAI

Upgrade the Snowflake Spark Connector to version 3.2.1 or later, which is the primary fix; consult the vendor release notes at https://docs.snowflake.com/en/release-notes/clients-drivers/spark-connector-2026 for the exact patched artifact and coordinates. Until you can upgrade, reduce exposure by not sharing a single Spark or Spark-SQL session across untrusted tenants or users (isolate workloads into per-tenant clusters), which removes the runtime-SET and staging-option injection paths at the cost of higher infrastructure overhead. Additionally, restrict who can supply connector options, staging options, and OAuth token-request URLs; run the connector's JDBC session under a least-privilege Snowflake role rather than a cluster-admin role to limit blast radius; and lock COPY operations to an allowlist of trusted, validated storage stages/locations to blunt redirection to attacker-controlled storage. These controls narrow but do not fully close the injection surface, so patching remains the definitive remedy.