Evidence used
- No CISA KEV confirmation is currently recorded.
- A structured source references public exploit or proof-of-concept material.
- EPSS is 0.45% for the current model date.
BlackTreeCVE IntelligenceFull vulnerability report · 2026
CVE-2026-94093High confidence
DLR-RM stable-baselines3 save_util.py VecNormalize.load deserialization
DLR-RM · stable-baselines3
5.3MediumCVSS 4.0
Within 7 days
Fix not verifiedMedium technical severity with public exploit material referenced by a structured source; prioritise exposed affected systems while verifying vendor guidance.
R
Open reassessment dashboard →Operational reassessment
Published severity in operational context
Published severityMedium→Operational priority:High, raised one band.upgradedsince 21 Sep 2026
Compensating controls
- Restrict the affected network interface to trusted sources where business-safe.
- Increase monitoring for the attack path and post-exploitation behaviour described in the report.
Verification
- Confirm that the asset runs DLR-RM stable-baselines3 and falls inside the recorded affected range.
- Recheck the vendor advisory before scheduling a change because no verified fixed version is currently retained.
- Validate exposure, authentication requirements and compensating controls in the actual environment.
- Reopen this reassessment when CVSS, KEV, EPSS, exploit evidence or remediation changes.
Medium technical severity with public exploit material referenced by a structured source; prioritise exposed affected systems while verifying vendor guidance.
Fix not verified01
What, why and how
A security vulnerability has been detected in DLR-RM stable-baselines3 up to 2.9.0. This affects the function PPO.load/load_replay_buffer/VecNormalize.load of the file save_util.py. Such manipulation leads to deserialization. It is possible to launch the attack remotely. The exploit has been disclosed publicly and may be used. In v2.9.0 the PyTorch tensor load path is hardened (weights_only=True), but that hardening was later reverted on master via PR #1913 "Hotfix: revert loading with weights_only=True" [blocked] to fix PyTorch 1.13 compat - so even the one "safe" path is inconsistent across versions. #2281 was closed as a duplicate of #1831 since both are unsafe pickle deserialization - but #1831's fix (PR #41) only gated the Hugging Face Hub loader in the separate huggingface_sb3 package. This finding covers the core stable_baselines3 load APIs (PPO.load, load_replay_buffer, VecNormalize.load), which have no safe mode or gate and remained exploitable in v2.9.0 until the outstanding hardening (PR #2264) ships.
What
A security vulnerability has been detected in DLR-RM stable-baselines3 up to 2.9.0. This affects the function PPO.load/load_replay_buffer/VecNormalize.load of the file save_util.py. Such manipulation leads to deserialization. It is possible to launch the attack remotely. The exploit has been disclosed publicly and may be used. In v2.9.0 the PyTorch tensor load path is hardened (weights_only=True), but that hardening was later reverted on master via PR #1913 "Hotfix: revert loading with weights_only=True" [blocked] to fix PyTorch 1.13 compat - so even the one "safe" path is inconsistent across versions. #2281 was closed as a duplicate of #1831 since both are unsafe pickle deserialization - but #1831's fix (PR #41) only gated the Hugging Face Hub loader in the separate huggingface_sb3 package. This finding covers the core stable_baselines3 load APIs (PPO.load, load_replay_buffer, VecNormalize.load), which have no safe mode or gate and remained exploitable in v2.9.0 until the outstanding hardening (PR #2264) ships.
Why
Attacker-influenced serialised data is reconstructed as trusted objects, which can invoke dangerous application behaviour.
How
An attacker operating through a network path may attempt exploitation when the stated preconditions are met. If successful, the issue may cause the confidentiality, integrity or availability impact described by the vendor.
What
A security vulnerability has been detected in DLR-RM stable-baselines3 up to 2.9.0. This affects the function PPO.load/load_replay_buffer/VecNormalize.load of the file save_util.py. Such manipulation leads to deserialization. It is possible to launch the attack remotely. The exploit has been disclosed publicly and may be used. In v2.9.0 the PyTorch tensor load path is hardened (weights_only=True), but that hardening was later reverted on master via PR #1913 "Hotfix: revert loading with weights_only=True" [blocked] to fix PyTorch 1.13 compat - so even the one "safe" path is inconsistent across versions. #2281 was closed as a duplicate of #1831 since both are unsafe pickle deserialization - but #1831's fix (PR #41) only gated the Hugging Face Hub loader in the separate huggingface_sb3 package. This finding covers the core stable_baselines3 load APIs (PPO.load, load_replay_buffer, VecNormalize.load), which have no safe mode or gate and remained exploitable in v2.9.0 until the outstanding hardening (PR #2264) ships.
Why
Attacker-influenced serialised data is reconstructed as trusted objects, which can invoke dangerous application behaviour.
How
An attacker operating through a network path may attempt exploitation when the stated preconditions are met. If successful, the issue may cause the confidentiality, integrity or availability impact described by the vendor.
02
Exploit reality and attack path
CVSS severity, EPSS forecast probability, public exploit material and CISA-confirmed exploitation are separate signals.
Observed exploitation No confirmed evidence
?
Confirmed exploitation and public exploit material are separate signals. Attacks can occur without public proof-of-concept or exploit code.No CISA KEV match was present at the last successful refresh. This means no confirmation from that source, not proof of no exploitation.
Public PoC / exploit material Reference recorded
?
Confirmed exploitation and public exploit material are separate signals. Attacks can occur without public proof-of-concept or exploit code.CISA Vulnrichment records proof-of-concept exploitation in its SSVC data. BlackTree has not independently executed or validated exploit material.
- Likely attack path
- a network path → Deserialization of Untrusted Data → cause the confidentiality, integrity or availability impact described by the vendor
- Attack surface
- Network
- Privileges required
- None: unauthenticated exploitation is possible
- User interaction
- Passive interaction required
- Attack complexity
- Low: no specialised conditions are recorded
- Security boundary
- Not a CVSS 4.0 base metric
- Weakness
?
CWE means Common Weakness Enumeration: a standard category for the underlying weakness. - CWE-502 ↗
CWE-502: Deserialization of Untrusted Data. The product deserializes untrusted data without sufficiently ensuring that the resulting data will be valid.
- CVSS vector
?
CVSS means Common Vulnerability Scoring System. The vector records the metric values used to calculate technical severity. CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:L/VI:L/VA:L/SC:N/SI:N/SA:N/E:PCommon Vulnerability Scoring System 4.0: the compact vector below is decoded into plain language.
AVNetworkAttack vector: The vulnerable component can be reached over a network.ACLowAttack complexity: No specialised conditions are required beyond attacker-controlled input.ATNoneAttack requirements: No additional deployment or execution condition is required.PRNonePrivileges required: The attacker does not need an account or existing privileges.UIPassiveUser interaction: A user must unknowingly interact with the vulnerable system.VCLowVulnerable-system confidentiality: A successful attack can cause a limited loss.VILowVulnerable-system integrity: A successful attack can cause a limited loss.VALowVulnerable-system availability: A successful attack can cause a limited loss.SCNoneSubsequent-system confidentiality: No direct loss is represented by this metric.SINoneSubsequent-system integrity: No direct loss is represented by this metric.SANoneSubsequent-system availability: No direct loss is represented by this metric.- Post-exploitation / living off the land
- No specific living-off-the-land technique is confirmed in the structured sources. Monitor normal administration tools for activity inconsistent with the affected service's baseline.
03
Patch and workaround
Operational remediation based on structured source evidence.
- Status
?
Patch availability is based on structured fixed-version fields and authoritative update references. If no fix is verified, check the vendor advisory before making a change. - Fix not verified
- Affected
- stable-baselines3: 2.0, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9.0
- Fixed
- No fixed version is explicitly recorded in the structured CVE data.
- Action
- No verified patch reference is present in the current structured sources. Check the vendor advisory before making a change.
- Workaround
- No verified workaround is recorded. If business-safe, reduce exposure to the affected interface and allow only trusted sources until authoritative guidance is available.
04
Evidence and provenance
Published 20 Sept 2026 · Last source change 21 Sept 2026, 16:21 UTC · CWE-502 · Deserialization of Untrusted Data
CVE recordCVE.org · 5.2
CVSS sourceCNA
EPSS source FIRST · tracked since 2026-09-21
?
The date BlackTree first stored a score for this CVE from the daily FIRST EPSS feed.European sourceENISA EUVD · EUVD-2026-83800
Product sourceCNA
Remediation sourceCVE/CNA references
CWE sourceCNA
NVD statusNVD not scheduled
Core structured fields are present and their contributing authorities are shown above.
Material change intelligence
View recent updates ↗What changed after publication
- Public exploit evidencePublic exploit evidence changed from No public exploit to Public exploit reference.CVE.org ↗
- Before
- No public exploit
- After
- Public exploit reference
- ENISA EUVD mappingEUVD-2026-83800 was added to the official ENISA EUVD mapping for this CVE.ENISA EUVD ↗
- Before
- not recorded
- After
- {"euvdId":"EUVD-2026-83800"}
- Catalogue recordCVE added to the BlackTree catalogue.CNA ↗
Technical terms and abbreviations used in this report
- CVE
- Common Vulnerabilities and Exposures: the public identifier for one disclosed vulnerability.
- CVSS
- Common Vulnerability Scoring System: a technical severity framework; it is not patching priority by itself.
- EPSS
- Exploit Prediction Scoring System: FIRST's estimate of the probability that exploitation activity will be observed in the next 30 days; it is a forecast, not confirmation.
- CWE
- Common Weakness Enumeration: the standard category describing the underlying software or hardware weakness.
- CNA
- CVE Numbering Authority: an organisation authorised to assign and publish CVE records.
- CISA ADP
- Cybersecurity and Infrastructure Security Agency Authorized Data Publisher: structured enrichment added to a CVE record.
- NVD
- National Vulnerability Database: NIST's enrichment service for CVE records.
- CERT / CSIRT
- A computer security incident response team that publishes warnings or coordinates incident response.
- PoC
- Proof of concept: public material that demonstrates or helps reproduce exploitation.
- CSAF
- Common Security Advisory Framework: a machine-readable format for security advisories.
- LoTL
- Living off the land: abuse of legitimate tools or system functions during an attack.