Tensorflow is vulnerable to integer overflows. The vulnerability exists due to a lack of input sanitization in the TfLiteIntArrayGetSizeInBytes
function of the common.c
file doesn’t control inputs which allows an attacker to craft a TFLite model causing an application crash.
github.com/tensorflow/tensorflow/blob/ca6f96b62ad84207fbec580404eaa7dd7403a550/tensorflow/lite/c/common.c#L24-L33
github.com/tensorflow/tensorflow/blob/ca6f96b62ad84207fbec580404eaa7dd7403a550/tensorflow/lite/c/common.c#L53-L60
github.com/tensorflow/tensorflow/commit/a1e1511dde36b3f8aa27a6ec630838e7ea40e091
github.com/tensorflow/tensorflow/security/advisories/GHSA-9gwq-6cwj-47h3