CVE-2020-15196
Heap buffer overflow in Tensorflow
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
Attack profile
How easy it is to exploit
- Attack vector
- Network
- Complexity
- High
- Privileges needed
- Low
- User interaction
- None
What an attacker can affect
- Confidentiality
- High
- Integrity
- High
- Availability
- High
Exploit likelihood
What each source reports
CVE List Score used
Heap buffer overflow in Tensorflow
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
CVE-2020-15196- CVSS 3.1
- 8.5 High
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H - Weakness
- CWE-119, CWE-122
- Updated
- 2026-11-02
EUVDEUVD-2020-0192
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
EUVD-2020-0192- CVSS 3.1
- 8.5 High
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H - Updated
- 2026-11-03
GitHub AdvisoriesGHSA-pg59-2f92-5cph
Heap buffer overflow in Tensorflow
Impact
The SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified:
https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/core/kernels/count_ops.cc#L110-L117
In the sparse and ragged count weights are still accessed in parallel with the data: https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/core/kernels/count_ops.cc#L199-L201
But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights.
Patches
We have patched the issue in 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and will release a patch release.
We recommend users to upgrade to TensorFlow 2.3.1.
For more information
Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.
Attribution
This vulnerability is a variant of GHSA-p5f8-gfw5-33w4
- Record
GHSA-pg59-2f92-5cph- CVSS 4.0
- 5.8 Medium
CVSS:4.0/AV:N/AC:L/AT:P/PR:L/UI:N/VC:N/VI:N/VA:N/SC:H/SI:H/SA:H - CVSS 3.1
- 8.5 High
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H - Weakness
- CWE-119, CWE-122, CWE-125
- Updated
- 2026-11-02
NVD
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
CVE-2020-15196- CVSS 3.1
- 8.5 High
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H - CVSS 3.1
- 9.9 Critical
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H - CVSS 2.0
- 6.5 Medium
AV:N/AC:L/Au:S/C:P/I:P/A:P - Weakness
- CWE-119, CWE-122, CWE-125
- Updated
- 2026-11-02
OSV
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
CVE-2020-15196- CVSS 3.1
- 9.9 Critical
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H - Updated
- 2026-11-02
OSVGHSA-pg59-2f92-5cph
Heap buffer overflow in Tensorflow
Impact
The SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified:
https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/core/kernels/count_ops.cc#L110-L117
In the sparse and ragged count weights are still accessed in parallel with the data: https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/core/kernels/count_ops.cc#L199-L201
But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights.
Patches
We have patched the issue in 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and will release a patch release.
We recommend users to upgrade to TensorFlow 2.3.1.
For more information
Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.
Attribution
This vulnerability is a variant of GHSA-p5f8-gfw5-33w4
- Record
GHSA-pg59-2f92-5cph- CVSS 4.0
- 5.8 Medium
CVSS:4.0/AV:N/AC:L/AT:P/PR:L/UI:N/VC:N/VI:N/VA:N/SC:H/SI:H/SA:H - CVSS 3.1
- 8.5 High
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H - Weakness
- CWE-119, CWE-122, CWE-125
- Updated
- 2026-11-02
Show 4 moreShow less
OSVBIT-tensorflow-2020-15196 No score
Heap buffer overflow in Tensorflow
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
BIT-tensorflow-2020-15196- Updated
- 2026-11-02
OSVPYSEC-2020-119 No score
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
PYSEC-2020-119- Updated
- 2026-11-02
OSVPYSEC-2020-276 No score
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
PYSEC-2020-276- Updated
- 2026-11-02
OSVPYSEC-2020-311 No score
In Tensorflow version 2.3.0, the SparseCountSparseOutput and RaggedCountSparseOutput implementations don't validate that the weights tensor has the same shape as the data. The check exists for DenseCountSparseOutput, where both tensors are fully specified. In the sparse and ragged count weights are still accessed in parallel with the data. But, since there is no validation, a user passing fewer weights than the values for the tensors can generate a read from outside the bounds of the heap buffer allocated for the weights. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
- Record
PYSEC-2020-311- Updated
- 2026-11-02
Affected products
Affected packages
| Package | Affected | Fixed |
|---|---|---|
tensorflowPyPI |
2.3.0 to 2.3.0 | 2.3.1 |
tensorflow-cpuPyPI |
2.3.0 to 2.3.0 | 2.3.1 |
tensorflow-gpuPyPI |
2.3.0 to 2.3.0 | 2.3.1 |
tensorflowBitnami |
2.3.0 up to 2.3.1 | 2.3.1 |
tensorflowPyPI |
2.3.0 up to 2.3.1 | 2.3.1 |
tensorflow-cpuPyPI |
2.3.0 up to 2.3.1 | 2.3.1 |
tensorflow-gpuPyPI |
2.3.0 up to 2.3.1 | 2.3.1 |
tensorflowPyPI |
before 2.3.1, plus listed versions 0.12.0rc00.12.0rc10.12.00.12.11.0.0 Show 45 moreShow less1.0.11.1.0rc01.1.0rc11.1.0rc21.1.01.2.0rc01.2.0rc11.2.0rc21.2.01.2.11.3.0rc01.3.0rc11.3.0rc21.3.01.4.0rc01.4.0rc11.4.01.4.11.5.0rc01.5.0rc11.5.01.5.11.6.0rc01.6.0rc11.6.01.7.0rc01.7.0rc11.7.01.7.11.8.0rc01.8.0rc11.8.01.9.0rc01.9.0rc11.9.0rc21.9.01.10.0rc01.10.0rc11.10.01.10.11.11.0rc01.11.0rc11.11.0rc21.11.01.12.0rc0 |
2.3.1 |
tensorflow-cpuPyPI |
before 2.3.1, plus listed versions 1.15.02.1.02.1.12.1.22.1.3 Show 6 moreShow less2.1.42.2.02.2.12.2.22.2.32.3.0 |
2.3.1 |
tensorflow-gpuPyPI |
before 2.3.1, plus listed versions 0.12.00.12.11.0.01.0.11.1.0 Show 41 moreShow less1.10.01.10.11.11.01.12.01.12.21.12.31.13.11.13.21.14.01.15.01.15.21.15.31.15.41.15.51.2.01.2.11.3.01.4.01.4.11.5.01.5.11.6.01.7.01.7.11.8.01.9.02.0.02.0.12.0.22.0.32.0.42.1.02.1.12.1.22.1.32.1.42.2.02.2.12.2.22.2.32.3.0 |
2.3.1 |
History
| Recorded | Changed | Sources |
|---|---|---|
| aliases | EUVD | |
| first indexed | CVE List, GitHub Advisories, NVD, OSV |
Dates are when Every CVE recorded each change. Upstream dates may differ.
In short
CVE-2020-15196 is a high vulnerability (CVSS 8.5) in Google Tensorflow, published September 25, 2020. No source lists public exploit code. EPSS puts the chance of exploitation in the next 30 days at 0.91%, the 60th percentile. Fixed in tensorflow 2.3.1.
References
github.com/tensorflow/tensorflow/releases/tag/v2.3.1
nvd.nist.gov/vuln/detail/CVE-2020-15196
Left out here: 54 versions. The JSON record has them.