Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAuthenticate as project contributor
-
DeliveryCraft malicious connection resource property
-
ExploitVictim's Studio Space startup interpolates value into shell
-
ExecutionExecute arbitrary commands in victim space
-
PersistHarvest member's temporary execution-role credentials
-
ImpactLateral movement / privilege escalation in project
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A05:2025 - Injection
ASVS 5.0
V1 - Encoding and Sanitization OS command sanitization, avoid shell exec
MITRE ATT&CK
2 techniques across 1 tactic: Execution The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires an authenticated remote user holding project contributor permissions within an Amazon SageMaker Unified Studio project (shared/multi-tenant project context), so this is an insider or compromised-collaborator scenario rather than external unauthenticated access. The attacker must supply a crafted connection resource property that is interpolated unsanitized into a shell invocation by the Studio Space startup validation script. The attack is additionally gated by user interaction (CVSS UI:P): a different project member's Studio Space must start/run that startup script for the injected command to execute against that member and expose their temporary execution-role credentials. Affected only on SageMaker Distribution 2.x<2.14.12, 3.x<3.9.12, 4.0.x<4.0.11, 4.1.x<4.1.11, 4.2.x<4.2.8, 4.3.x<4.3.5, 4.4.x<4.4.3; end-of-support minor lines receive no fix and must be migrated. |
| Impact | This is a genuinely high-priority flaw, not merely a high-CVSS paper risk. The CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:L/UI:P/VC:H/VI:H/VA:H/SC:H/SI:H) scores 9.3 and the metrics are internally consistent with the description: network reachable, low attack complexity, low privileges (an authenticated project contributor), and passive user interaction (a victim project member must use/restart their Studio Space). The CWE-78 OS command injection yields full code execution in another member's Studio Space plus theft of that member's temporary execution-role credentials, which explains the high confidentiality/integrity/availability and the subsequent-system (SC:H/SI:H) impact - a credential-theft/lateral-movement primitive within a shared project. Tempering factors: exploitation is NOT unauthenticated (PR:L requires valid project contributor permissions, i.e. an insider or compromised collaborator account) and requires victim interaction (UI:P). There is no CISA KEV listing, no public POC, and no EPSS score in the provided data, so there is no evidence of active or widespread exploitation - the risk today is the capability, not observed campaigns. Net: credible insider/lateral-escalation threat warranting prompt patching in multi-tenant SageMaker Unified Studio projects, but gated behind existing project access and user interaction rather than being an internet-facing unauthenticated RCE. |
| Exploit ScenarioDerived | A user holding project contributor permissions in an Amazon SageMaker Unified Studio project crafts a malicious connection resource whose property value contains shell metacharacters. When a different project member's Studio Space runs the startup validation script, the unsanitized property is interpolated into a shell invocation, executing the attacker's commands in that member's space and exposing their temporary execution-role credentials. No public exploit code is identified at time of analysis, and the attack depends on the victim's Studio Space starting (UI:P). |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 02, 2026
OS command injection in the Studio Space startup validation script of Amazon SageMaker Distribution (2.x, 3.x, and 4.x lines, as used by Amazon SageMaker Unified Studio) allows an authenticated remote user with project contributor permissions to execute arbitrary commands inside another project member's Studio Space and steal that member's temporary execution-role credentials via a crafted connection resource property. Exploitation is a shared/multi-tenant insider or compromised-collaborator scenario rather than unauthenticated external access, and it is additionally gated by user interaction because the victim's Studio Space must start and run the vulnerable startup script for the injected command to reach that member. No public exploit identified at time of analysis; vendor-released patches exist for supported lines, and end-of-support minor lines will not receive a fix.
RemediationAI
Vendor-released patch: upgrade to 2.14.12, 3.9.12, 4.0.11, 4.1.11, 4.2.8, 4.3.5, or 4.4.3 as applicable to the minor line in use (see https://github.com/aws/sagemaker-distribution/releases and https://aws.amazon.com/security/security-bulletins/2026-125-aws/). For Amazon SageMaker Unified Studio, Studio Spaces automatically adopt the latest patch of their minor line on restart once patched images are deployed, so no manual version selection is required - restart spaces to pick up the fix. Users on minor lines that have reached end of support must move to a supported minor line, since no patched release will be issued for those lines. Until spaces are restarted onto a patched image, compensating controls include restricting project contributor permissions to trusted members and auditing/removing untrusted connection resource properties, at the cost of reduced collaboration flexibility and the risk that un-reviewed properties remain; note that these controls limit but do not eliminate the shared-project trust boundary, because any contributor can still influence configuration consumed by another member's space, and they do not stop credential exposure when the victim's startup script runs.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
The root cause is CWE-78 (Improper Neutralization of Special Elements used in an OS Command), arising where a connection resource property is interpolated into a shell invocation without sanitization inside the SageMaker Distribution Studio Space startup validation script. Amazon SageMaker Unified Studio organizes work into projects whose members each operate an isolated Studio Space that assumes a temporary execution role; because those spaces share a project-level configuration surface, a crafted connection property supplied by one contributor is evaluated in the context of another member's space when that member's startup script runs. The escalation is therefore a cross-boundary command execution and credential-exfiltration issue within a shared multi-tenant environment, rated CVSS:4.0 9.3 by the vendor with PR:L and UI:P, and independently assessed as CVSS:3.1 AV:N/AC:L/PR:L/UI:R/S:C/C:H/I:H/A:H - consistent on privileged-but-authenticated access, network reachability, required user interaction, changed scope, and high confidentiality/integrity/availability impact. The CPE cpe:2.3:a:aws:sagemaker-distribution identifies the affected component as the SageMaker Distribution image rather than the Unified Studio control plane itself.
Know when Sagemaker Distribution gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:P/VC:H/VI:H/VA:H/SC:H/SI:H/SA:L
Network-reachable injection needing only project-contributor rights (PR:L) and victim interaction (UI:R); impact crosses into another member's space and credentials, so scope changes (S:C) with high C/I/A.
Primary rating from Vendor (AMZN).
CVSS VectorVendor: AMZN
Lifecycle Timeline
2DescriptionCVE.org
OS command injection in the Studio Space startup validation script in Amazon SageMaker Distribution 2.x before 2.14.12, 3.x before 3.9.12, 4.0.x before 4.0.11, 4.1.x before 4.1.11, 4.2.x before 4.2.8, 4.3.x before 4.3.5, and 4.4.x before 4.4.3, as used by Amazon SageMaker Unified Studio, might allow an authenticated remote user with project contributor permissions to execute arbitrary commands in another project member's Studio Space and obtain that member's temporary execution role credentials via a crafted connection resource property that is interpolated into a shell invocation without neutralization.
To remediate this issue, users should upgrade to version 2.14.12, 3.9.12, 4.0.11, 4.1.11, 4.2.8, 4.3.5, or 4.4.3, as applicable to the minor line in use. Users on minor lines that have reached end of support must move to a supported minor line, because no patched version will be released for those lines. In Amazon SageMaker Unified Studio, Studio Spaces adopt the latest patch of their minor line on restart once the patched images are deployed, so no version selection is required.
Technical POC Analysis
The vulnerable code is the Studio Space startup validation script in Amazon SageMaker Distribution images.
Full technical POC analysis - root cause, step-by-step exploitation and patch breakdown - is part of the Pro and Business plans.
Executive ViewAI
Amazon SageMaker Unified Studio contains a critical security flaw that allows a project contributor to execute commands in another member's workspace and steal temporary execution-role credentials, potentially exposing sensitive data and resources across projects. This shared/multi-tenant insider risk affects organizations using SageMaker Unified Studio for collaborative machine learning, and leadership should prioritize remediation to maintain tenant isolation and protect intellectual property. Currently, no public exploit identified at time of analysis, and patch available per vendor advisory for supported lines, though end-of-support minor lines will not receive a fix.
Apply vendor patch.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
OS command injection occurs when an application passes unsanitized user input directly into system shell commands.
EUVD-2026-91795
NVD
CVE.org
EPSS
Vulnrichment
GitHub
Exploit-DB