CVSS 3.1
7.5 highCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:NHigh (7.5) in langchain langchain core. EPSS 1.5% estimated probability of exploitation activity in the next 30 days. Fixed in langchain-core 1.2.22.
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:Nlangchain-core (PyPI): fixed in 1.2.22Explain CVE-2026-34070 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-34070?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-34070?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-34070 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.
From the NVD's CPE matches.
| Product | Affected |
|---|---|
| langchain:langchain_core | before 1.2.22 |
| Advisory | Ecosystem | Package | Affected | Fixed in |
|---|---|---|---|---|
| GHSA-qh6h-p6c9-ff54 | PyPI | langchain-core | <1.2.22 | 1.2.22 |
| PYSEC-2026-2193 | PyPI | langchain-core | <1.2.22 | 1.2.22 |
Red Hat rates this Important.
This flaw is exploitable in applications that accept prompt configs from untrusted sources, including low-code AI builders and API wrappers that expose `load_prompt_from_config()`.
Also, the affected functions (`load_prompt`, `load_prompt_from_config` and the `.save()` method on prompt classes) are undocumented legacy APIs. They are superseded by the `dumpd`/`dumps`/`load`/`loads` serialization APIs in `langchain_core.load`, which do not perform filesystem reads and use an allowlist-based security model.
An attacker who controls or influences the prompt configuration dictionary can read files outside the intended directory, such as cloud-mounted secrets, internal system prompts, cloud credentials, Kubernetes manifests, CI/CD configs and application settings.
Due to these reasons, this vulnerability has been rated with an important severity.
As described in the statement section, the vulnerable methods are legacy APIs and their use should be avoided. To mitigate this issue, the dumpd, dumps, load and loads methods from langchain_core.load should be used, as they supersede the legacy API and provide a more secure serialization model.
| Product | Package | State | Advisory |
|---|---|---|---|
| Red Hat Ansible Automation Platform 2.5 | ansible-automation-platform-25/lightspeed-rhel8:1780082949 | Fixed 2026-06-09 | RHSA-2026:24766 |
| Red Hat Ansible Automation Platform 2.5 | ansible-automation-platform-25/lightspeed-rhel8:1787229385 | Fixed 2026-08-24 | RHSA-2026:59159 |
| Red Hat OpenShift AI 3.3 | rhoai/odh-llama-stack-core-rhel9:1782471587 | Fixed 2026-07-09 | RHSA-2026:37275 |
| Red Hat OpenShift AI 3.3 | rhoai/odh-trustyai-nemo-guardrails-server-rhel9:1782471606 | Fixed 2026-07-09 | RHSA-2026:37275 |
| OpenShift Lightspeed | openshift-lightspeed/lightspeed-service-api-rhel9 | Affected | |
| Red Hat Ansible Automation Platform 2 | ansible-automation-platform-24/lightspeed-rhel8 | Will not fix | |
| Red Hat Ansible Automation Platform 2 | ansible-automation-platform-25/lightspeed-chatbot-rhel8 | Will not fix | |
| Red Hat Ansible Automation Platform 2 | ansible-automation-platform-26/lightspeed-rhel9 | Will not fix | |
| Red Hat OpenShift AI (RHOAI) | rhoai/odh-mlflow-rhel9 | Not affected |
Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-34070.
Repositories on GitHub named after this CVE. Nobody has checked them and they are not scanned: some hold a working proof of concept, some nothing useful, and some malware made to look like one. Read the code before you run any of it.
The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits.