keras-team keras up to 3.15.0 TorchModuleWrapper TorchModuleWrapper.from_config deserialization

Summaryinfo

A vulnerability marked as critical has been reported in keras-team keras up to 3.15.0. This affects the function TorchModuleWrapper.from_config of the component TorchModuleWrapper. The manipulation leads to deserialization. This vulnerability is documented as CVE-2026-12484. The attack can be initiated remotely. There is not any exploit available.

Detailsinfo

A vulnerability, which was classified as critical, has been found in keras-team keras up to 3.15.0. This issue affects the function TorchModuleWrapper.from_config of the component TorchModuleWrapper. The manipulation with an unknown input leads to a deserialization vulnerability. Using CWE to declare the problem leads to CWE-502. The product deserializes untrusted data without sufficiently verifying that the resulting data will be valid. Impacted is confidentiality, integrity, and availability. The summary by CVE is:

A vulnerability in keras-team/keras version 3.15.0 allows unsafe deserialization of attacker-controlled PyTorch pickle data through the public `keras.layers.TorchModuleWrapper.from_config` method. This method invokes `torch.load(..., weights_only=False)` without requiring an explicit unsafe opt-in, such as a `safe_mode=False` parameter. When called outside a `SafeModeScope(True)` context, the absence of an ambient safe mode state permits unsafe deserialization by default. This issue can lead to arbitrary code execution if untrusted Keras layer configurations are processed using this method. The vulnerability arises because the method does not enforce safe deserialization practices unless explicitly guarded by Keras safe mode.

It is possible to read the advisory at huntr.com. The identification of this vulnerability is CVE-2026-12484 since 06/17/2026. The exploitation is known to be easy. The attack may be initiated remotely. No form of authentication is needed for a successful exploitation. Technical details of the vulnerability are known, but there is no available exploit. The pricing for an exploit might be around USD $0-$5k at the moment (estimation calculated on 07/20/2026).

There is no information about possible countermeasures known. It may be suggested to replace the affected object with an alternative product.

The vulnerability is also documented in the vulnerability database at EUVD (EUVD-2026-45872). Statistical analysis made it clear that VulDB provides the best quality for vulnerability data.

Productinfo

Vendor

• keras-team

Name

• keras

Version

• 3.0
• 3.1
• 3.2
• 3.3
• 3.4
• 3.5
• 3.6
• 3.7
• 3.8
• 3.9
• 3.10
• 3.11
• 3.12
• 3.13
• 3.14
• 3.15.0

CPE 2.3info

• 🔒
• 🔒
• 🔒

CPE 2.2info

• 🔒
• 🔒
• 🔒

CVSSv4info

CVSSv3info

CVSSv2info

Exploitinginfo

Threat Intelligenceinfo

Countermeasuresinfo

Timelineinfo

Sourcesinfo

Entryinfo

Statistical analysis made it clear that VulDB provides the best quality for vulnerability data.

Discussion