CVE-2026-90614: FedML-AI FedML MQTT+S3 Communication Backend remote_storage.py S3Storage.read_model deserialization
A weakness has been identified in FedML-AI FedML up to 0.9.6. Affected by this issue is the function S3Storage.readmodel of the file fedml/core/distributed/communication/s3/remotestorage.py of the component MQTT+S3 Communication Backend. This manipulation of the argument s3keystr causes deserialization. Remote exploitation of the attack is possible. The project was informed of the problem early through an issue report but has not responded yet.
Affected Software
Event History
Frequently Asked Questions
Which deployments are in scope?
FedML-AI FedML versions up to 0.9.6 are affected when using the MQTT+S3 Communication Backend and its S3Storage.read_model function.
What access does an attacker need to exploit this issue?
The CVSS vector indicates that exploitation is network-accessible, requires low privileges, and does not require user interaction. The issue involves manipulation of the s3_key_str argument supplied to S3Storage.read_model.
What is the expected impact if exploitation succeeds?
The reported CVSS assessment rates confidentiality, integrity, and availability impact as low. The vulnerability is categorized as an input-validation weakness involving deserialization.