Security readout for executives and security teams
Plain-English summary
CVE-2021-37665 is a TensorFlow input-validation flaw in MKL requantization code. A low-privileged local attacker may trigger undefined behavior or out-of-bounds heap access, potentially affecting confidentiality, integrity, and availability. It matters most on shared ML systems where users can run or influence TensorFlow workloads.
Executive priority
Treat as high priority for shared or production ML environments using affected TensorFlow releases. The issue is not reported as actively exploited, but the impact rating is high and fixed releases are available, so remediation should be scheduled promptly through normal dependency update channels.
Technical view
Affected TensorFlow versions fail to fully validate input tensor dimensions and arguments in MKL requantization per-channel operations. The advisory describes null pointer reference binding and heap out-of-bounds access. CVSS 3.1 is 7.8, local attack vector, low complexity, low privileges, no user interaction, high CIA impact.
Likely exposure
Exposure is likely limited to systems running affected TensorFlow versions with relevant MKL requantization paths reachable by local users or jobs. Shared notebooks, training hosts, and multi-user ML pipelines deserve priority review. Evidence does not support treating this as a general remote network vulnerability.
Exploitation context
The source bundle does not show CISA KEV listing or active exploitation evidence. The CVSS vector indicates local, low-privileged exploitation without user interaction. Practical risk depends on whether an attacker can run or influence TensorFlow code and tensors on the affected system.
Researcher notes
The cited issue is CWE-20 incomplete input validation in TensorFlow MKL requantization implementations. Sources name commits 9e62869465573cb2d9b5053f1fa02a81fce21d69 and 203214568f5bc237603dbab6e1fd389f1572f5c9 as fixes. Avoid assuming exploitability beyond local low-privileged execution without additional environment evidence.
Mitigation direction
- Upgrade TensorFlow to 2.6.0 or fixed supported branch releases.
- Use TensorFlow 2.5.1, 2.4.3, or 2.3.4 where branch pinning is required.
- Prioritize shared ML systems that allow untrusted local TensorFlow jobs.
- Restrict local job execution to trusted users until affected environments are patched.
- Check the TensorFlow advisory and commits for branch-specific guidance.
Validation and detection
- Inventory TensorFlow versions in hosts, notebooks, containers, and lockfiles.
- Flag versions >=2.5.0 <2.5.1, >=2.4.0 <2.4.3, and <2.3.4.
- Identify workloads using TensorFlow MKL or Intel-optimized execution paths.
- Confirm upgraded images and dependency locks no longer include affected versions.
- Run existing ML and security regression tests after the TensorFlow update.
Public sources used
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
CWE-20: Exact CWE lookup
Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
Open ATT&CK lookupCVE-2021-37665 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
Open ATT&CK lookup- Severity
- High
- CVSS
- 7.8 (3.1)
- Known Exploited
- No
- Published
Vector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
CNA and ADP enrichment extracted from CVE v5
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.
CVSS vector scores
1 official scoreWe collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H1.85.9Primary CVE scoreVulnerability scoring details
Base CVSS 3.1 score
7.8HighVector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Source materials
- CVE List V5 sourceCVE List V5
- https://github.com/tensorflow/tensorflow/security/advisories/GHSA-v82p-hv3v-p6qpCVE reference · x_refsource_CONFIRM
- https://github.com/tensorflow/tensorflow/commit/203214568f5bc237603dbab6e1fd389f1572f5c9CVE reference · x_refsource_MISC
- https://github.com/tensorflow/tensorflow/commit/9e62869465573cb2d9b5053f1fa02a81fce21d69CVE reference · x_refsource_MISC
Products and packages named in the record
CWE details
CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.
Improper Input Validation
Improper Input Validation represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.