PYSEC-2026-3223 Undefined behavior in `SparseTensorSliceDataset`

Impact The implementation of SparseTensorSliceDataset has an undefined behavior: under certain condition it can be made to dereference a nullptr value: python import tensorflow as tf import numpy as np tf.rawops.SparseTensorSliceDataset indices=, values=, denseshape=1,1 The 3 input arguments...

CVE-2026-54601

FastGPT is an open source AI knowledge base platform. From 4.14.17 to before 4.15.0-beta4, FastGPT allows an authenticated tenant user to call POST /api/core/dataset/collection/create/reTrainingCollection in a way that persists a server-owned datasetId value from another tenant. This creates mixe...

CVE-2026-55418

FastGPT is an open source AI knowledge base platform. Prior to v4.15.0-beta5, two FastGPT file handlers authorize an unrelated resource and then sign or read an S3 object using a key taken directly from the request, without checking that the key belongs to the caller's team. Because S3 object key...

CVE-2026-54601 FastGPT: reTrainingCollection allows server-owned datasetId override causing cross-tenant authorization confusion

FastGPT is an open source AI knowledge base platform. From 4.14.17 to before 4.15.0-beta4, FastGPT allows an authenticated tenant user to call POST /api/core/dataset/collection/create/reTrainingCollection in a way that persists a server-owned datasetId value from another tenant. This creates mixe...

CVE-2026-54601 FastGPT: reTrainingCollection allows server-owned datasetId override causing cross-tenant authorization confusion

FastGPT is an open source AI knowledge base platform. From 4.14.17 to before 4.15.0-beta4, FastGPT allows an authenticated tenant user to call POST /api/core/dataset/collection/create/reTrainingCollection in a way that persists a server-owned datasetId value from another tenant. This creates mixe...

CVE-2026-54601

FastGPT is an open source AI knowledge base platform. From 4.14.17 to before 4.15.0-beta4, FastGPT allows an authenticated tenant user to call POST /api/core/dataset/collection/create/reTrainingCollection in a way that persists a server-owned datasetId value from another tenant. This creates mixe...

CVE-2026-54601

CVE-2026-54601 affects FastGPT up to 4.15.0-beta4, where an authenticated tenant user can abuse POST /api/core/dataset/collection/create/reTrainingCollection to persist a server-owned datasetId from another tenant. This yields mixed dataset objects and downstream endpoints (dataset, collection, a...

CVE-2026-50530

DataEase prior to version 2.10.24 exposes a vulnerability in the share mode chart data interface. The API path POST /de2api/chartData/getData only validates that sceneId matches the resourceId in the link token and does not verify whether tableId and field IDs in the request body belong to the sh...

CVE-2026-50530

DataEase is an open source data visualization and analysis tool. Prior to 2.10.24, a share mode chart data interface only validates that sceneId matches the resourceId in the link token and fails to validate whether tableId and field IDs in the request body belong to the shared resource, allowing...

PYSEC-2026-1308 docarray prototype pollution

A vulnerability was found in docarray up to 0.40.1. It has been rated as critical. Affected by this issue is the function getitem of the file /docarray/data/torchdataset.py of the component Web API. The manipulation leads to improperly controlled modification of object prototype attributes...

PYSEC-2026-1406 GluonCV Arbitrary File Write via TarSlip

A vulnerability in the ImageClassificationDataset.fromcsv API of the dmlc/gluon-cv repository, version 0.10.0, allows for arbitrary file write. The function downloads and extracts tar.gz files from URLs without proper sanitization, making it susceptible to a TarSlip vulnerability. Attackers can...

PYSEC-2026-1628 MindsDB Cross-site Scripting vulnerability

A cross-site scripting XSS vulnerability exists in all versions of the MindsDB platform, enabling the execution of a JavaScript payload whenever a user enumerates an ML Engine, database, project, or dataset containing arbitrary JavaScript code within the web UI...

PYSEC-2026-1191 Apache Superset: Improper data authorization when creating a new dataset

Apache Superset with custom roles that include can write on dataset and without all data access permissions, allows for users to create virtual datasets to data they don't have access to. These users could then use those virtual datasets to get access to unauthorized data. This issue affects Apac...

PYSEC-2026-1249 Out of memory error when submitting the dataset form with a specially-crafted field

Impact When submitting a POST request to the /dataset/new endpoint including either the auth cookie or the Authorization header with a specially-crafted field, an attacker can create an out-of-memory error in the hosting server. To trigger this error the user needs to have permissions to create o...

PYSEC-2026-1179 Apache Superset Server-Side Request Forgery vulnerability

A malicious actor who has been authenticated and granted specific permissions in Apache Superset may use the import dataset feature in order to conduct Server-Side Request Forgery attacks and query internal resources on behalf of the server where Superset is deployed. This vulnerability exists in...

Arbitrary HDF5 File Read via Virtual Dataset Bypass in keras-team/keras

...

PT-2026-56259

Name of the Vulnerable Software and Affected Versions FastGPT versions 4.14.17 through 4.15.0-beta3 Description An authenticated tenant user can call the 'POST /api/core/dataset/collection/create/reTrainingCollection' endpoint to persist a datasetId value belonging to another tenant. This results...

PYSEC-2026-776 Apache Superset allows authenticated users to access metadata they have no permission to

Apache Superset up to 1.5.1 allowed for authenticated users to access metadata information related to datasets they have no permission on. This metadata included the dataset name, columns and metrics...

CVE-2026-12480

A flaw was found in Keras. An attacker can craft a malicious model archive or weights file containing a Virtual Dataset VDS that references external files on a victim's system. When a user loads this malicious model, the external file is transparently read. This vulnerability leads to information...

GHSA-26C4-7VV6-867J Keras: HDF5 virtual datasets can disclose local files

Keras versions up to and including 3.13.2 are vulnerable to an arbitrary HDF5 file read due to an incomplete fix for CVE-2026-1669. The vulnerability resides in the H5IOStore.verifydataset and fileeditor.py methods, which fail to check the dataset.isvirtual property of HDF5 datasets. This allows ...