EUVD-2025-210786
GHSA-ppjh-2wmr-qh8x
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessHost malicious cross-origin page
-
DeliveryLure authenticated North user to visit
-
ExploitBrowser sends credentialed cross-origin request
-
ExecutionServer omits Origin validation
-
PersistScript reads API response
-
ImpactExfiltrate sensitive AI data
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A02:2025 - Security Misconfiguration
ASVS 5.0
V13 - Configuration CORS policy restriction
|
| Exploitation Conditions | Exploitation requires that the North server be configured (as shipped in v1.1.5) to return permissive CORS headers and fail to validate the Origin of incoming requests - that misconfiguration IS the condition. Practically limiting factors: the victim must have an active authenticated session to North and must be tricked into visiting or rendering attacker-controlled web content (browser-driven, so effectively UI:R), and the target data must be reachable via credentialed cross-origin requests. It is not a serverside-only unauthenticated exploit despite the PR:N/UI:N metrics in the published vector; the attacker cannot read data unless a real authenticated user's browser makes the request. |
| Impact | The signals conflict. The published CVSS:3.1 vector (AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H = 9.8) rates this critical with full CIA loss and no user interaction, but that is inconsistent with the vulnerability class: a CORS/Origin-validation flaw is only exploitable when a logged-in victim visits an attacker-controlled page, so realistic exploitation requires user interaction (UI:R) and primarily yields confidentiality loss (information disclosure per the tags), not availability impact. The A:H and UI:N metrics appear overstated - verify with vendor. Supporting a lower real-world priority: EPSS is 0.14% (4th percentile), there is no KEV listing and no identified exploit code - only a third-party GitHub writeup. Net: treat as a genuine information-disclosure bug worth patching, not a wormable RCE despite the 9.8 headline. There is no SSVC decision data in the input. |
| Exploit ScenarioDerived | An attacker hosts a malicious web page and lures a user who is currently authenticated to Cohere North into visiting it. The page's JavaScript issues credentialed cross-origin requests to the North API; because the server does not validate the Origin, the browser permits the attacker's script to read the responses, exfiltrating conversations, documents, or API data. A public technical write-up describing the flaw exists on GitHub, but no weaponized exploit code was identified. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Aug 27, 2026
Sensitive-data exposure in Cohere North AI v1.1.5 stems from the server accepting cross-origin requests without validating the Origin header, letting any untrusted website read authenticated API responses in a victim's browser. Affects the North enterprise AI platform; tagged as CORS misconfiguration / information disclosure. No public exploit or active exploitation is identified, and EPSS is low (0.14%, 4th percentile), though the NVD-published CVSS of 9.8 likely overstates real-world impact.
RemediationAI
No vendor-released patch version is identified at time of analysis, so contact Cohere for a fixed North build and monitor https://cohere.com/north for updates. As concrete compensating controls: configure the server (or a fronting reverse proxy) to reject or not reflect unrecognized Origin headers and replace any wildcard/reflected Access-Control-Allow-Origin with a strict allowlist of trusted first-party origins - side effect: legitimate integrations on other domains will break until explicitly allowlisted. Avoid pairing Access-Control-Allow-Credentials: true with a permissive origin. Where feasible, place North behind an authenticating gateway or restrict access to a trusted network/VPN so browsers of authenticated users cannot be lured into cross-origin requests from arbitrary internet sites; trade-off is reduced accessibility for remote/SaaS-style users. Refer to the GitHub advisory for the specific misconfigured endpoint details.
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
The root cause is CWE-942 (Permissive Cross-Domain Policy with Untrusted Domains). Cohere North is a self/enterprise-hosted generative-AI workspace exposing HTTP/JSON APIs consumed by a browser front end. Browsers enforce the Same-Origin Policy, but servers can relax it via CORS response headers (Access-Control-Allow-Origin / Access-Control-Allow-Credentials). Here the server reflects or wildcards the requesting Origin instead of matching it against an allowlist, so a page on any attacker-controlled domain can issue credentialed cross-origin XHR/fetch calls and read the responses. CPE data is uninformative (cpe:2.3:a:n/a:n/a:*), so exact affected components beyond 'North AI 1.1.5' cannot be enumerated from the feed.
Know when Cors Misconfiguration 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
Sources disagree (Medium–Critical)AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
CORS flaw needs a logged-in victim to visit attacker content, so UI:R; primarily discloses data (C:H) with limited state-change (I:L) and no availability impact (A:N).
vuln.today treats the vendor’s rating as authoritative. A higher third-party CVSS (e.g. CISA-ADP) is shown for transparency but does not drive the headline severity.
CVSS VectorVendor: mitre
Lifecycle Timeline
3DescriptionCVE.org
Cohere North AI v1.1.5 was discovered to contain excessively permissive cross-domain policy with untrusted domains. This occurs via the server failing to validate the Origin header of incoming connection requests.
Executive ViewAI
Cohere North AI v1.1.5 contains a CORS misconfiguration allowing unauthorized browser-based disclosure of sensitive API responses, affecting all organizations running this version. This exposes enterprise data including business intelligence and user information to attackers capable of deceiving users into visiting malicious websites.
Without an available patch, implement Web Application Firewall rules to enforce strict Origin header validation and restrict cross-origin requests to internal domains only; disable cross-origin access entirely if external API consumption is not required; isolate affected systems behind network segmentation to limit attacker reach; enforce API authentication and rate limiting as additional control layers; and monitor API access logs for anomalous cross-origin request patterns.
Threat Intelligence Overview
NVD
CVE.org
EPSS
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
Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.
More in Cors Misconfiguration
View allCross-origin data theft in LightRAG server versions prior to 1.5.4 allows any malicious website to make authenticated, c
Casdoor is a UI-first Identity and Access Management (IAM) / Single-Sign-On (SSO) platform. Rated high severity (CVSS 8.
In Eclipse Theia 0.3.9 to 1.8.1, the "mini-browser" extension allows a user to preview HTML files in an iframe inside th
Unauthenticated cross-origin RCE in Windows ML CLI (winml-cli < 0.4.0) is reachable when a developer runs the `serve` co
Bruno is an open source IDE for exploring and testing APIs. Rated high severity (CVSS 8.7), this vulnerability is no aut
memos is a privacy-first, lightweight note-taking service. Rated high severity (CVSS 8.1), this vulnerability is remotel
Cross-origin read bypass in Electron's custom protocol handler exposes sensitive application data to remote pages via fe
Information disclosure in Progress MOVEit Transfer (all versions before 2025.1.5, and 2026.0.0 before 2026.0.3) stems fr
SCG Policy Manager, all versions, contains an overly permissive Cross-Origin Resource Policy (CORP) vulnerability. Rated
In Directus before 9.7.0, the default settings of CORS_ORIGIN and CORS_ENABLED are true. Rated critical severity (CVSS 9
A malicious website could have learned the size of a cross-origin resource that supported Range requests. Rated critical
Remote code execution in SiYuan desktop application (versions prior to 3.6.2) allows unauthenticated remote attackers to
Exploit-DB