CVE-2020-15266: In Tensorflow before version 2.4.0, when the `boxes` argument of `tf.image.crop_and_resize` has a very large value, the
lowvulnerability
security
Summary
TensorFlow versions before 2.4.0 have a bug in the `tf.image.crop_and_resize` function where very large values in the `boxes` argument are converted to NaN (a special floating point value meaning "not a number"), causing undefined behavior and a segmentation fault (a crash from illegal memory access). This vulnerability affects the CPU implementation of the function.
Solution / Mitigation
Upgrade to TensorFlow version 2.4.0 or later, which contains the patch. TensorFlow nightly packages (development builds) after commit eccb7ec454e6617738554a255d77f08e60ee0808 also have the issue resolved.
Vulnerability Details
CVSS Score
3.7(low)
EPSS (30-day exploit probability)
EPSS: 0.1%
Classification
Attack SophisticationModerate
Impact (CIA+S)
availability
AI Component TargetedFramework
Affected Vendors
Original source: https://nvd.nist.gov/vuln/detail/CVE-2020-15266
First tracked: February 15, 2026 at 08:38 PM
Classified by LLM (prompt v3) · confidence: 95%