EUVD-2025-210784
GHSA-gpwh-5jwg-6gc2
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessReach North batch_upload API
-
DeliverySubmit crafted malicious file
-
ExploitBypass file-type validation (CWE-434)
-
ExecutionFile stored in executable path
-
PersistBackend executes payload
-
ImpactFull host compromise
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A06:2025 - Insecure Design
ASVS 5.0
V5 - File Handling File type validation, storage isolation
MITRE ATT&CK
2 techniques across 2 tactics: Command and Control, Persistence The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires reaching the /v1/my_drive/batch_upload API and submitting a crafted file that bypasses upload validation, targeting Cohere North AI v1.1.5. The 'my_drive' scoping strongly suggests a valid authenticated user session is needed to access the personal drive upload function, which would limit exploitation to those who can obtain or provision an account - a meaningful limiting factor that contradicts the PR:N in the supplied CVSS vector and must be confirmed with the vendor. The supplied vector otherwise indicates network reachability with low complexity and no user interaction; no non-default configuration is documented as a prerequisite, but affected-version data beyond 1.1.5 is unconfirmed. |
| Impact | The supplied CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H = 9.8 Critical) implies trivial, unauthenticated, network-based RCE. However, the endpoint name /v1/my_drive/batch_upload strongly implies a per-user, session-scoped drive that typically requires an authenticated account to reach - this conflicts with the PR:N in the input vector and should be verified with the vendor, since it materially changes real-world risk. EPSS is only 0.15% (5th percentile) and there is no CISA KEV listing, indicating no confirmed active or widespread exploitation. The signals therefore diverge: raw CVSS says maximum urgency, but exploitation probability is low and authentication may be a genuine barrier. Net assessment: a serious flaw (file-upload-to-RCE) warranting prompt remediation for any org running North, but not clearly a mass-exploited internet-wide emergency on current evidence. |
| Exploit ScenarioDerived | An attacker who can reach the North API sends a crafted request to /v1/my_drive/batch_upload containing a malicious file (for example a server-interpretable script) disguised as a normal document; because the file type and content are not properly validated, the payload is stored and then executed by the backend, giving the attacker code execution on the host. A public GitHub write-up documents the issue, lowering the effort to reproduce it, though no packaged exploit is confirmed in the feeds. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Aug 27, 2026
Remote code execution in Cohere North AI v1.1.5 arises from an arbitrary file upload flaw in the /v1/my_drive/batch_upload endpoint, where insufficient validation of uploaded file type/content lets an attacker place a malicious file that the server subsequently executes. The affected component is part of Cohere's North enterprise AI knowledge platform, and successful exploitation yields full compromise of confidentiality, integrity, and availability. No public exploit is identified in the intelligence feeds, though a third-party technical write-up exists on GitHub; EPSS is low (0.15%, 5th percentile), indicating limited observed exploitation activity to date.
RemediationAI
No vendor-released patch identified at time of analysis; no fixed version is present in the intelligence data, so contact Cohere for a fixed build of North beyond 1.1.5 and monitor http://cohere.com, http://north.com, and the NVD entry (https://nvd.nist.gov/vuln/detail/CVE-2025-61165) for an advisory. As compensating controls, restrict network access to the /v1/my_drive/batch_upload endpoint to trusted, authenticated users via a reverse proxy or WAF rule, and enforce strict server-side upload validation - allowlist permitted extensions and MIME types, store uploads outside any web or executable path, strip execute permissions, and rename files to non-executable names; the trade-off is that tightening or blocking batch_upload may break legitimate document-ingestion workflows, so validate against expected file types before deploying. Review upload directories and application logs for anomalous or executable files as a detective control.
Affected VersionsENISA
Affects your stack
Find out whether the stacks and servers you run are on an affected version.
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
This is a CWE-434 (Unrestricted Upload of File with Dangerous Type) weakness. The vulnerable path is the /v1/my_drive/batch_upload API of Cohere North, an enterprise generative-AI assistant and knowledge-management platform that ingests user documents into a personal 'drive.' Because batch upload accepts attacker-controlled files without adequately restricting extension, MIME type, or content, a crafted file (e.g., a server-side script or a payload interpreted by the backend runtime) can be written to a location where the application later executes it, converting a file-ingestion feature into arbitrary code execution. The single CPE provided (cpe:2.3:a:n/a:n/a:*) is a placeholder with no real vendor/product tokens, so affected-product enumeration must come from the description rather than the CPE.
Know when N A 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
AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Network-reachable file upload yielding RCE (C/I/A High); PR:L rather than PR:N because the per-user 'my_drive' endpoint most likely requires an authenticated account.
Primary rating from Vendor (mitre).
CVSS VectorVendor: mitre
Lifecycle Timeline
3DescriptionCVE.org
An arbitrary file upload vulnerability in the /v1/my_drive/batch_upload component of cohere North AI v1.1.5 allows attackers to exeute arbitrary code via uploading a crafted file.
Executive ViewAI
Cohere North AI v1.1.5 contains a critical remote code execution vulnerability in its file upload service that enables attackers to execute arbitrary code on affected systems. This directly threatens organizations relying on this enterprise AI platform, risking exposure of proprietary datasets, customer information, and critical infrastructure. With no security patch currently available from the vendor, immediate risk mitigation is essential despite currently limited observed exploitation activity.
No vendor-released patch identified at time of analysis. Interim mitigations include: disable or restrict the /v1/my_drive/batch_upload endpoint to authorized internal systems only, implement Web Application Firewall rules to block or rate-limit file uploads to this endpoint, enforce strict file type and content validation at the application layer, isolate affected Cohere North AI instances to segregated network segments to contain potential compromise, and implement comprehensive logging and alerting on all file upload attempts for rapid detection of exploitation attempts.
Threat Intelligence Overview
NVD
CVE.org
EPSS
Vulnrichment
VulDB
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
Malicious file upload attacks exploit insufficient validation when web applications accept user-provided files.
Remote Code Execution represents the critical moment when an attacker successfully runs arbitrary code on a target system without physical access.
LibreNMS before 24.10.0 allows a remote attacker to execute arbitrary code via OS command injection involving AboutContr
Reachable assertion in Open5GS up to version 2.7.6 allows remote unauthenticated attackers to crash the Diameter CER han
A Command Injection issue in the payload build page in BYOB (Build Your Own Botnet) 2.0 allows attackers to execute arbi
Denial of service in Open5GS 2.7.6 and earlier allows network-adjacent attackers with low-privilege Diameter Gx peer acc
Denial of service in Open5GS HSS (Home Subscriber Server) versions 2.7.0 through 2.7.6 is triggered remotely by manipula
Remote command injection in Control Web Panel allows unauthenticated attackers to execute arbitrary OS commands as root
LINQPad before 5.52.01 Pro edition is vulnerable to Unsafe Deserialization in LINQPad.AutoRefManager::PopulateFromCache(
Remote code execution in dash-uploader (Python package for Plotly Dash) versions 0.1.0 through 0.7.0a2 allows unauthenti
Command injection in LiteLLM 1.18.10 lets attackers who can configure MCP (Model Context Protocol) servers supply arbitr
SQL injection in AzuraCast's backup restore flow (versions up to and including 0.23.7) lets an attacker achieve privileg
Arbitrary file read in Yu AI Code Mother v4.3 lets anonymous, remote attackers retrieve files outside the intended previ
Remote code execution in APScheduler (all versions through 3.10.x and 4.0.0a5) is achievable when applications deseriali
Exploit-DB