Langchain: Improper Neutralization of Special Elements in Output Used by a Downstream Component
CVE-2023-32786 is a high-severity improper neutralization of special elements in output used by a downstream component vulnerability in Langchain, with a CVSS 3.1 base score of 7.5 out of 10. It can be exploited remotely over the network by an unauthenticated attacker, without any action from a user. Affected Langchain versions: 0.0.155 and earlier. The record does not name a fixed version, so check the vendor advisory. There is no confirmed exploitation so far. EPSS estimates a 0.71% chance of exploitation within 30 days, higher than 54% of scored CVEs.
High severity with no evidence of exploitation so far.
What to do
- Check which version of Langchain you run against the affected versions below, and install the vendor's latest release.
- Until you can update, restrict network access to the affected service so it is not reachable from untrusted networks.
Affected products and versions
Version data comes from NVD's product matching. Confirm against the vendor advisory before deciding you are unaffected.
How the severity score breaks down
- Attack vector
- NetworkCan be exploited remotely over a network, including the internet.
- Attack complexity
- LowNo special conditions are needed; the attack works reliably.
- Privileges required
- NoneNo account or login is needed.
- User interaction
- NoneNo action by a victim is required.
- Scope
- UnchangedImpact stays within the vulnerable component.
- Confidentiality
- HighAll data handled by the component can be read.
- Integrity
- NoneNothing can be modified.
- Availability
- NoneAvailability is not affected.
Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:N. Scored by NVD.
Official description
In Langchain through 0.0.155, prompt injection allows an attacker to force the service to retrieve data from an arbitrary URL, essentially providing SSRF and potentially injecting content into downstream tasks.
Source: the CVE record as published by its numbering authority, via NVD and CVE.org.
The underlying weakness: Improper Neutralization of Special Elements in Output Used by a Downstream Component
The product constructs all or part of a command, data structure, or record using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify how it is parsed or interpreted when it is sent to a downstream component.
How developers prevent this class of flaw
- Programming languages and supporting technologies might be chosen which are not subject to these issues.
- Utilize an appropriate mix of allowlist and denylist parsing to filter control-plane syntax from all input.
Timeline
- CVE record published.
- CISA published its SSVC assessment (exploitation: none, automatable: yes, technical impact: partial).
- Record last changed materially (score, affected products, status or references).
References
Third-party and government advisories
- gist.github.com/rharang/d265f46fc3161b31ac2e81db44d662e1Third Party Advisory
Other vulnerabilities in Langchain
- 10.0 Critical severity
- 9.8 Critical severity
- 7.8 High severity
- 7.7 High severity
- 8.1 High severity
- 9.8 Critical severity
Recent Improper Neutralization of Special Elements in Output Used by a Downstream Component vulnerabilities
- 7.5 High severity
- 8.8 High severity
- 8.8 High severity
- 8.8 High severity
- 7.5 High severity
- 8.8 High severity
Spotted an error? Tell us.