CVE-2026-1462 (GCVE-0-2026-1462)

Vulnerability from cvelistv5 – Published: 2026-04-13 14:55 – Updated: 2026-04-13 18:53
VLAI?
Title
Safe Mode Bypass in keras-team/keras
Summary
A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of `.keras` models, even when `safe_mode=True`. This bypasses the security guarantees of `safe_mode` and enables arbitrary attacker-controlled code execution during model inference under the victim's privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the `from_config()` method.
Severity ?
8.8 (High)
CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
CWE
  • CWE-502 - Deserialization of Untrusted Data
Assigner
References
Impacted products
Vendor Product Version
keras-team keras-team/keras Affected: unspecified , < 3.13.2 (custom)
Create a notification for this product.
Show details on NVD website
To clipboard 

{
  "containers": {
    "adp": [
      {
        "metrics": [
          {
            "other": {
              "content": {
                "id": "CVE-2026-1462",
                "options": [
                  {
                    "Exploitation": "poc"
                  },
                  {
                    "Automatable": "no"
                  },
                  {
                    "Technical Impact": "total"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2026-04-13T18:53:01.058651Z",
                "version": "2.0.3"
              },
              "type": "ssvc"
            }
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2026-04-13T18:53:12.291Z",
          "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
          "shortName": "CISA-ADP"
        },
        "title": "CISA ADP Vulnrichment"
      }
    ],
    "cna": {
      "affected": [
        {
          "product": "keras-team/keras",
          "vendor": "keras-team",
          "versions": [
            {
              "lessThan": "3.13.2",
              "status": "affected",
              "version": "unspecified",
              "versionType": "custom"
            }
          ]
        }
      ],
      "descriptions": [
        {
          "lang": "en",
          "value": "A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of `.keras` models, even when `safe_mode=True`. This bypasses the security guarantees of `safe_mode` and enables arbitrary attacker-controlled code execution during model inference under the victim\u0027s privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the `from_config()` method."
        }
      ],
      "metrics": [
        {
          "cvssV3_0": {
            "attackComplexity": "LOW",
            "attackVector": "NETWORK",
            "availabilityImpact": "HIGH",
            "baseScore": 8.8,
            "baseSeverity": "HIGH",
            "confidentialityImpact": "HIGH",
            "integrityImpact": "HIGH",
            "privilegesRequired": "NONE",
            "scope": "UNCHANGED",
            "userInteraction": "REQUIRED",
            "vectorString": "CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
            "version": "3.0"
          }
        }
      ],
      "problemTypes": [
        {
          "descriptions": [
            {
              "cweId": "CWE-502",
              "description": "CWE-502 Deserialization of Untrusted Data",
              "lang": "en",
              "type": "CWE"
            }
          ]
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2026-04-13T14:55:28.649Z",
        "orgId": "c09c270a-b464-47c1-9133-acb35b22c19a",
        "shortName": "@huntr_ai"
      },
      "references": [
        {
          "url": "https://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331c"
        },
        {
          "url": "https://github.com/keras-team/keras/commit/b6773d3decaef1b05d8e794458e148cb362f163f"
        }
      ],
      "source": {
        "advisory": "7e78d6f1-6977-4300-b595-e81bdbda331c",
        "discovery": "EXTERNAL"
      },
      "title": "Safe Mode Bypass in keras-team/keras"
    }
  },
  "cveMetadata": {
    "assignerOrgId": "c09c270a-b464-47c1-9133-acb35b22c19a",
    "assignerShortName": "@huntr_ai",
    "cveId": "CVE-2026-1462",
    "datePublished": "2026-04-13T14:55:28.649Z",
    "dateReserved": "2026-01-27T04:14:51.848Z",
    "dateUpdated": "2026-04-13T18:53:12.291Z",
    "state": "PUBLISHED"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.2",
  "vulnerability-lookup:meta": {
    "epss": {
      "cve": "CVE-2026-1462",
      "date": "2026-04-14",
      "epss": "0.0006",
      "percentile": "0.18735"
    },
    "nvd": "{\"cve\":{\"id\":\"CVE-2026-1462\",\"sourceIdentifier\":\"security@huntr.dev\",\"published\":\"2026-04-13T15:17:18.967\",\"lastModified\":\"2026-04-13T15:17:18.967\",\"vulnStatus\":\"Received\",\"cveTags\":[],\"descriptions\":[{\"lang\":\"en\",\"value\":\"A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of `.keras` models, even when `safe_mode=True`. This bypasses the security guarantees of `safe_mode` and enables arbitrary attacker-controlled code execution during model inference under the victim\u0027s privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the `from_config()` method.\"}],\"metrics\":{\"cvssMetricV30\":[{\"source\":\"security@huntr.dev\",\"type\":\"Secondary\",\"cvssData\":{\"version\":\"3.0\",\"vectorString\":\"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H\",\"baseScore\":8.8,\"baseSeverity\":\"HIGH\",\"attackVector\":\"NETWORK\",\"attackComplexity\":\"LOW\",\"privilegesRequired\":\"NONE\",\"userInteraction\":\"REQUIRED\",\"scope\":\"UNCHANGED\",\"confidentialityImpact\":\"HIGH\",\"integrityImpact\":\"HIGH\",\"availabilityImpact\":\"HIGH\"},\"exploitabilityScore\":2.8,\"impactScore\":5.9}]},\"weaknesses\":[{\"source\":\"security@huntr.dev\",\"type\":\"Primary\",\"description\":[{\"lang\":\"en\",\"value\":\"CWE-502\"}]}],\"references\":[{\"url\":\"https://github.com/keras-team/keras/commit/b6773d3decaef1b05d8e794458e148cb362f163f\",\"source\":\"security@huntr.dev\"},{\"url\":\"https://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331c\",\"source\":\"security@huntr.dev\"}]}}",
    "vulnrichment": {
      "containers": "{\"adp\": [{\"title\": \"CISA ADP Vulnrichment\", \"metrics\": [{\"other\": {\"type\": \"ssvc\", \"content\": {\"id\": \"CVE-2026-1462\", \"role\": \"CISA Coordinator\", \"options\": [{\"Exploitation\": \"poc\"}, {\"Automatable\": \"no\"}, {\"Technical Impact\": \"total\"}], \"version\": \"2.0.3\", \"timestamp\": \"2026-04-13T18:53:01.058651Z\"}}}], \"providerMetadata\": {\"orgId\": \"134c704f-9b21-4f2e-91b3-4a467353bcc0\", \"shortName\": \"CISA-ADP\", \"dateUpdated\": \"2026-04-13T18:53:07.966Z\"}}], \"cna\": {\"title\": \"Safe Mode Bypass in keras-team/keras\", \"source\": {\"advisory\": \"7e78d6f1-6977-4300-b595-e81bdbda331c\", \"discovery\": \"EXTERNAL\"}, \"metrics\": [{\"cvssV3_0\": {\"scope\": \"UNCHANGED\", \"version\": \"3.0\", \"baseScore\": 8.8, \"attackVector\": \"NETWORK\", \"baseSeverity\": \"HIGH\", \"vectorString\": \"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H\", \"integrityImpact\": \"HIGH\", \"userInteraction\": \"REQUIRED\", \"attackComplexity\": \"LOW\", \"availabilityImpact\": \"HIGH\", \"privilegesRequired\": \"NONE\", \"confidentialityImpact\": \"HIGH\"}}], \"affected\": [{\"vendor\": \"keras-team\", \"product\": \"keras-team/keras\", \"versions\": [{\"status\": \"affected\", \"version\": \"unspecified\", \"lessThan\": \"3.13.2\", \"versionType\": \"custom\"}]}], \"references\": [{\"url\": \"https://huntr.com/bounties/7e78d6f1-6977-4300-b595-e81bdbda331c\"}, {\"url\": \"https://github.com/keras-team/keras/commit/b6773d3decaef1b05d8e794458e148cb362f163f\"}], \"descriptions\": [{\"lang\": \"en\", \"value\": \"A vulnerability in the `TFSMLayer` class of the `keras` package, version 3.13.0, allows attacker-controlled TensorFlow SavedModels to be loaded during deserialization of `.keras` models, even when `safe_mode=True`. This bypasses the security guarantees of `safe_mode` and enables arbitrary attacker-controlled code execution during model inference under the victim\u0027s privileges. The issue arises due to the unconditional loading of external SavedModels, serialization of attacker-controlled file paths, and the lack of validation in the `from_config()` method.\"}], \"problemTypes\": [{\"descriptions\": [{\"lang\": \"en\", \"type\": \"CWE\", \"cweId\": \"CWE-502\", \"description\": \"CWE-502 Deserialization of Untrusted Data\"}]}], \"providerMetadata\": {\"orgId\": \"c09c270a-b464-47c1-9133-acb35b22c19a\", \"shortName\": \"@huntr_ai\", \"dateUpdated\": \"2026-04-13T14:55:28.649Z\"}}}",
      "cveMetadata": "{\"cveId\": \"CVE-2026-1462\", \"state\": \"PUBLISHED\", \"dateUpdated\": \"2026-04-13T18:53:12.291Z\", \"dateReserved\": \"2026-01-27T04:14:51.848Z\", \"assignerOrgId\": \"c09c270a-b464-47c1-9133-acb35b22c19a\", \"datePublished\": \"2026-04-13T14:55:28.649Z\", \"assignerShortName\": \"@huntr_ai\"}",
      "dataType": "CVE_RECORD",
      "dataVersion": "5.2"
    }
  }
}


Log in or create an account to share your comment.




Tags
Taxonomy of the tags.




Sightings

Author Source Type Date

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.



Detection rules are retrieved from Rulezet.