Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessGain adjacent-network position
-
DeliveryMITM Ollama update channel
-
ExploitServe malicious update executable
-
ExecutionVerification stub returns success
-
PersistSilent auto-update stages payload
-
ImpactExecute arbitrary code on host
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A08:2025 - Software or Data Integrity Failures
ASVS 5.0
V15 - Secure Coding and Architecture Code integrity verification
|
| Exploitation Conditions | Exploitation requires the attacker to control or tamper with the update executable Ollama for Windows downloads - in practice an adjacent-network man-in-the-middle position on the update channel or a compromised/impersonated update source (reflected in AV:A and AT:P, the 'attack requirement present' flag). The target must be running Ollama's Windows build in the confirmed range 0.12.10-0.17.5 with automatic updates active (the default, silent behavior). No authentication and no user interaction are needed once the attacker can supply the payload (PR:N/UI:N), and the silent update means the victim receives no warning. Limiting factors: the flaw is Windows-only (macOS/Linux perform the skipped check), the attacker cannot exploit purely remotely from an arbitrary Internet location without a network/update-channel foothold, and reachability depends on the client actually performing an update check while the attacker holds position. |
| Impact | Signals here are mixed and should temper the 7.7 CVSS 4.0 score. The vector (AV:A/AC:L/AT:P) requires an adjacent-network position and a present attack requirement - the attacker must be able to tamper with or impersonate the update download channel (e.g., man-in-the-middle on the network path or a compromised update endpoint), not merely send a packet from anywhere on the Internet. That materially constrains real-world exploitability, which is consistent with the very low EPSS (0.01%, 1st percentile) and CISA SSVC marking Exploitation as 'none' and Automatable as 'no'. Countervailing factors are severe: technical impact is 'total', the update runs silently with no user interaction (PR:N/UI:N), and the flaw affects the trust anchor of the whole update process, so a successful attack yields full host compromise. Net: not a mass-exploitation, drop-everything issue, but a high-severity supply-chain/update-integrity weakness that becomes serious for anyone on untrusted networks or facing a capable adjacent attacker. The 'Information Disclosure' tag understates the true impact, which is arbitrary code execution. |
| Exploit ScenarioDerived | An attacker who shares an adjacent network with a target (e.g., the same corporate LAN, VPN segment, or public Wi-Fi) positions themselves as a man-in-the-middle on Ollama's update traffic and serves a malicious executable when the Windows client checks for updates. Because the Windows verification routine always returns success, the tampered payload passes the integrity check and Ollama's silent auto-update stages and executes it without any user prompt, giving the attacker code execution on the host. No public exploit code is known, but the attack is straightforward for anyone already holding the required network position. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Jul 24, 2026
Silent auto-update code execution in Ollama for Windows (versions 0.12.10 through 0.17.5) stems from an update-verification routine that unconditionally returns success, so no signature or trust check is performed before staging and executing downloaded update payloads (CWE-494). Because Ollama on Windows applies updates silently and automatically, an attacker who can substitute the update executable can have arbitrary code installed and run without user awareness. No public exploit identified at time of analysis; EPSS is very low (0.01%) and CISA SSVC records exploitation as none, though technical impact is rated total.
RemediationAI
No vendor-released patch identified at time of analysis - CERT-PL reports the maintainers were notified early but did not respond with a fix or an authoritative vulnerable-version range, so no fixed version can be cited. Until a signed release confirming the verification routine is restored on Windows is published, compensating controls are required: disable or block Ollama's automatic update mechanism where the platform allows, and prevent the client from silently fetching updates by restricting outbound access from Ollama hosts to only trusted, TLS-pinned update endpoints (trade-off: manual updates must then be applied and verified out-of-band). Ensure hosts run Ollama only on trusted networks and enforce network-layer protections (e.g., prevent ARP/DNS spoofing, require validated TLS) to remove the adjacent man-in-the-middle position the attack needs (trade-off: does not fix the underlying missing check). Monitor the Ollama install directory for unexpected executable changes as a detection control. Track https://cert.pl/en/posts/2026/04/CVE-2026-42248/ and https://ollama.com/ for a patched build and upgrade as soon as one is available.
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
Ollama is a local large-language-model runtime that ships a built-in auto-update mechanism. On non-Windows platforms the client validates the digital signature/authenticity of the downloaded installer before running it, but the Windows implementation of that verification function is effectively a stub that always returns success, so any bytes delivered as the 'update' are accepted. This is a textbook CWE-494 (Download of Code Without Integrity Check): the trust decision that should gate execution has been removed on one platform. The affected component is the Ollama Windows client itself (CPE cpe:2.3:a:ollama:ollama), and the impact is code execution in the security context of the updating process.
Know when Ollama 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:A/AC:L/AT:P/PR:N/UI:N/VC:H/VI:H/VA:H/SC:L/SI:L/SA:L
AV:A and AC:H because exploitation needs an adjacent man-in-the-middle position on the update channel; PR:N/UI:N due to silent unattended updates, with full C/I/A from code execution.
Primary rating from NVD.
CVSS VectorNVD
Lifecycle Timeline
2DescriptionCVE.org
Ollama for Windows does not perform integrity or authenticity verification of downloaded update executables. Unlike other platforms, the Windows implementation of the update verification routine unconditionally returns success so no digital signature or trust validation is performed before staging or executing update payloads, enabling attacker‑supplied executables to be accepted and later executed by the application.
Critically, Ollama for Windows performs silent automatic updates, so the malicious payload may be installed automatically without user awareness.
Maintainers of this project were notified early about this vulnerability, but didn't respond with the details of vulnerability or vulnerable version range. Versions from 0.12.10 to 0.17.5 were tested and confirmed as vulnerable, other versions were not tested but might also be vulnerable.
Executive ViewAI
Ollama for Windows (versions 0.12.10-0.17.5) contains a silent auto-update vulnerability (CVE-2026-42248) that executes downloaded code without signature verification, enabling arbitrary code execution without user awareness. Windows users running affected versions face total compromise risk if network or supply-chain attackers intercept updates; the automatic, invisible execution model makes this a critical infrastructure integrity threat.
No vendor-released patch identified at time of analysis. Interim mitigations: (1) Disable auto-update functionality in Ollama configuration immediately; (2) Block outbound connections from Ollama to update servers using firewall rules; (3) Run Ollama in isolated network segments or containers with restricted egress; (4) Downgrade to version 0.12.9 or earlier if operationally feasible; (5) Deploy endpoint detection and response (EDR) with behavioral monitoring for Ollama process anomalies or unexpected child process execution; (6) Monitor Windows process logs for Ollama-related activity.
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
Information disclosure occurs when an application unintentionally exposes sensitive data that aids attackers in reconnaissance or directly compromises security.
Ollama before 0.1.34 does not validate the format of the digest (sha256 with 64 hex digits) when getting the model path,
Remote code execution in Ollama for Windows (versions 0.12.10 through 0.17.5) lets an attacker who can influence HTTP up
Remote denial of service in Ollama (HEAD f0078ae) allows unauthenticated attackers to crash the entire server by uploadi
Denial of service in Ollama's multi-modal image processing (versions 0.11.5-rc0 through 0.13.5) lets remote attackers cr
An issue was discovered in Ollama before 0.1.46. Rated high severity (CVSS 8.2), this vulnerability is remotely exploita
A vulnerability in the Ollama server version 0.5.11 allows a malicious user to cause a Denial of Service (DoS) attack by
Information disclosure in Ollama's model quantization engine (confirmed affected in v0.13.5) allows remote, unauthentica
A vulnerability in ollama/ollama versions <=0.3.14 allows a malicious user to upload and create a customized GGUF model
An issue in ollama v.0.12.10 allows a remote attacker to cause a denial of service via the fs/ggml/gguf.go, function rea
An issue in ollama v.0.12.10 allows a remote attacker to cause a denial of service via the GGUF decoder [CVSS 7.5 HIGH]
A divide by zero vulnerability exists in ollama/ollama version v0.3.3. Rated high severity (CVSS 7.5), this vulnerabilit
A vulnerability in ollama/ollama versions <=0.3.14 allows a malicious user to create a customized GGUF model file that,
EUVD-2026-26210
NVD
CVE.org
EPSS
Vulnrichment
VulDB
GitHub PoC
Exploit-DB