GHSA-v638-38fc-rhfv — aws-encryption-sdk
MEDIUMGHSA-v638-38fc-rhfv is a medium-severity (CVSS 4.7) CWE-757 vulnerability in aws-encryption-sdk. A fix is available for aws-encryption-sdk — see the affected versions and patch details below.
AWS Encryption SDK for Python: Key commitment policy bypass via shared key cache
Exploitation Status
No confirmed exploitation observed yet
- CISA’s own triage has not observed active exploitation or public proof-of-concept code for this CVE as of its last assessment.
Exploitation and automatability from CISA’s SSVC triage for GHSA-v638-38fc-rhfv.
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
How urgent is this, really
GHSA-v638-38fc-rhfv plotted by exploitation likelihood (EPSS) against impact (CVSS). The shaded corner — EPSS 50%+ and CVSS 7.0+ — is where this CVE doesn't sit, though severity or exploitability alone can still warrant action.
Where this sits among everything scored
Of 379,145 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Real counts from FIRST.org, not a sample — log-scaled since the landscape is heavily right-skewed.
Real-World Exposure
aws-encryption-sdk🐍aws-encryption-sdkReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects PyPI packages — download data is not available via public APIs for these ecosystems.
Description
Summary
AWS Encryption SDK (ESDK) for Python is a client-side encryption library. An issue exists where, under certain circumstances, a specific cryptographic algorithm downgrade in the caching layer might allow an authenticated local threat actor to bypass key commitment policy enforcement via a shared key cache, resulting in ciphertext that can be decrypted to multiple different plaintexts.
Impact
This issue requires all of the following conditions to be true: (1) Two ESDK for Python clients with different commitment policies share a single CachingCryptoMaterialsManager instance within the same process. (2) The client with the weaker commitment policy encrypts first, warming the cache. (3) Both clients use matching encryption contexts. (4) Both clients use the pre-configured default algorithm suite.
These conditions may occur during a migration from ESDK for Python v1 to newer versions, as v1 did not support key commitment.
When the weaker-policy client encrypts first, the cache stores encryption materials that do not enforce key commitment. Subsequent callers — including those configured to require key commitment — are served these cached materials instead of generating new ones that satisfy their policy. This results in encryption without key commitment, meaning the same ciphertext can be validly decrypted to different plaintexts under different keys (the "Invisible Salamanders" issue; see https://github.com/google/security-research/security/advisories/GHSA-wqgp-vphw-hphf). A threat actor who controls ciphertext can cause a recipient to decrypt a message different from what the sender encrypted, breaking message integrity.
Impacted versions
- From 2.0 to 2.5.1
- From 3.0 to 3.3.0
- From 4.0 to 4.0.4
Patches
This issue has been addressed in ESDK for Python versions 3.3.1 and 4.0.5. We recommend upgrading to the latest version and ensuring any forked or derivative code is patched to incorporate the new fixes.
Workarounds
If a customer requires operating multiple instances of the Python ESDK each with differently configured key commitment policies, they must not share a key cache.
References If there are any questions or comments about this advisory, contact AWS Security through the vulnerability reporting page or directly via email to [email protected]. Please do not create a public GitHub issue.
Acknowledgement
Thanks to 1seal.org for collaborating on this issue through the coordinated vulnerability disclosure process.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐍PyPI | aws-encryption-sdk | ≥ 2.0.0&&< 3.3.1 | 3.3.1pip install --upgrade 'aws-encryption-sdk==3.3.1' |
| 🐍PyPI | aws-encryption-sdk | ≥ 4.0.0&&< 4.0.5 | 4.0.5pip install --upgrade 'aws-encryption-sdk==4.0.5' |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for aws-encryption-sdk, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update aws-encryption-sdk to 3.3.1 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-v638-38fc-rhfv is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 Security's impact-aware SCA analyses which vulnerable code paths your application actually calls, so a match like GHSA-v638-38fc-rhfv can be triaged on real exposure rather than presence alone.
Tailored to GHSA-v638-38fc-rhfv. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is GHSA-v638-38fc-rhfv in your dependencies?
O3 Security finds GHSA-v638-38fc-rhfv across PyPI dependencies, including transitive ones, and its impact-aware SCA ranks findings by whether your code actually calls the vulnerable path.