{"id":"CVE-2026-12481","aliases":["CLSA-2026-1790928189","DEBIAN-CVE-2026-12481","EUVD-2026-41600","GHSA-5gwj-m78q-7pq3","PYSEC-2026-3631","UBUNTU-CVE-2026-12481"],"status":"published","title":"Deserialization of Untrusted Data in keras-team/keras","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","severity":{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-09","score":0.00719,"percentile":0.54156,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":true},"cwes":["CWE-502"],"published":"2026-07-03T20:36:05Z","modified":"2026-10-28T02:30:59Z","fixes":[{"ecosystem":"PyPI","package":"keras","fixed_in":["3.12.3","3.15.0"],"sources":["ghsa","osv"]}],"affected":[{"source":"ghsa","ecosystem":"PyPI","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"fixed":"3.12.3"}]}]},{"source":"ghsa","ecosystem":"PyPI","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"3.13.0"},{"fixed":"3.15.0"}]}]},{"source":"osv","ecosystem":"TuxCare:PyPI","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"2.15.0"},{"fixed":"2.15.0.post3+tuxcare"}]}],"versions":["2.15.0","2.15.0.post1+tuxcare","2.15.0.post2+tuxcare"],"bundled":4},{"source":"osv","ecosystem":"Debian:11","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"}]}],"versions":["2.3.1+dfsg-3","2.3.1+dfsg2-1"]},{"source":"osv","ecosystem":"PyPI","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"fixed":"3.12.3"}]}],"versions":["0.2.0","0.3.0","0.3.1","0.3.2","0.3.3","1.0.0","1.0.1","1.0.2","1.0.3","1.0.4","1.0.5","1.0.6","1.0.7","1.0.8","1.1.0","1.1.1","1.1.2","1.2.0","1.2.1","1.2.2","2.0.0","2.0.1","2.0.2","2.0.3","2.0.4","2.0.5","2.0.6","2.0.7","2.0.8","2.0.9","2.1.0","2.1.1","2.1.2","2.1.3","2.1.4","2.1.5","2.1.6","2.10.0","2.10.0rc0","2.10.0rc1","2.11.0","2.11.0rc0","2.11.0rc1","2.11.0rc2","2.11.0rc3","2.12.0","2.12.0rc0","2.12.0rc1","2.13.1","2.13.1rc0","2.13.1rc1","2.14.0","2.14.0rc0","2.15.0","2.15.0rc0","2.15.0rc1","2.2.0","2.2.1","2.2.2","2.2.3","2.2.4","2.2.5","2.3.0","2.3.1","2.4.0","2.4.1","2.4.2","2.4.3","2.5.0rc0","2.6.0","2.6.0rc0","2.6.0rc1","2.6.0rc2","2.6.0rc3","2.7.0","2.7.0rc0","2.7.0rc2","2.8.0","2.8.0rc0","2.8.0rc1","2.9.0","2.9.0rc0","2.9.0rc1","2.9.0rc2","3.0.0","3.0.1","3.0.2","3.0.3","3.0.4","3.0.5","3.1.0","3.1.1","3.10.0","3.11.0","3.11.1","3.11.2","3.11.3","3.12.0","3.12.1","3.12.2","3.2.0","3.2.1","3.3.0","3.3.1","3.3.2","3.3.3","3.4.0","3.4.1","3.5.0","3.6.0","3.7.0","3.8.0","3.9.0","3.9.1","3.9.2"]},{"source":"osv","ecosystem":"PyPI","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"3.13.0"},{"fixed":"3.15.0"}]}],"versions":["3.13.0","3.13.1","3.13.2","3.14.0","3.14.1"]},{"source":"osv","ecosystem":"PyPI","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"fixed":"3.12.3"},{"introduced":"3.13.0"},{"fixed":"3.15.0"}]}],"versions":["0.2.0","0.3.0","0.3.1","0.3.2","0.3.3","1.0.0","1.0.1","1.0.2","1.0.3","1.0.4","1.0.5","1.0.6","1.0.7","1.0.8","1.1.0","1.1.1","1.1.2","1.2.0","1.2.1","1.2.2","2.0.0","2.0.1","2.0.2","2.0.3","2.0.4","2.0.5","2.0.6","2.0.7","2.0.8","2.0.9","2.1.0","2.1.1","2.1.2","2.1.3","2.1.4","2.1.5","2.1.6","2.10.0","2.10.0rc0","2.10.0rc1","2.11.0","2.11.0rc0","2.11.0rc1","2.11.0rc2","2.11.0rc3","2.12.0","2.12.0rc0","2.12.0rc1","2.13.1","2.13.1rc0","2.13.1rc1","2.14.0","2.14.0rc0","2.15.0","2.15.0rc0","2.15.0rc1","2.2.0","2.2.1","2.2.2","2.2.3","2.2.4","2.2.5","2.3.0","2.3.1","2.4.0","2.4.1","2.4.2","2.4.3","2.5.0rc0","2.6.0","2.6.0rc0","2.6.0rc1","2.6.0rc2","2.6.0rc3","2.7.0","2.7.0rc0","2.7.0rc2","2.8.0","2.8.0rc0","2.8.0rc1","2.9.0","2.9.0rc0","2.9.0rc1","2.9.0rc2","3.0.0","3.0.1","3.0.2","3.0.3","3.0.4","3.0.5","3.1.0","3.1.1","3.10.0","3.11.0","3.11.1","3.11.2","3.11.3","3.12.0","3.12.1","3.12.2","3.13.0","3.13.1","3.13.2","3.14.0","3.14.1","3.2.0","3.2.1","3.3.0","3.3.1","3.3.2","3.3.3","3.4.0","3.4.1","3.5.0","3.6.0","3.7.0","3.8.0","3.9.0","3.9.1","3.9.2"]},{"source":"osv","ecosystem":"Ubuntu:18.04:LTS","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"}]}],"versions":["1.0.7-2","2.1.1-1"]},{"source":"osv","ecosystem":"Ubuntu:20.04:LTS","package":"keras","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"}]}],"versions":["2.2.4-1"]}],"products":[{"vendor":"keras","product":"keras"},{"vendor":"keras-team","product":"keras-team/keras"}],"references":[{"url":"https://huntr.com/bounties/59ceaed1-c8a3-4135-8f94-169ade02823d","tags":["exploit","Exploit","Third Party Advisory","web","report"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-12481","tags":["advisory"]},{"url":"https://github.com/keras-team/keras/pull/23048","tags":["web"]},{"url":"https://github.com/keras-team/keras/pull/23165","tags":["web"]},{"url":"https://github.com/keras-team/keras/commit/55888d3becbdbb45dc16a55b489900f911e2dde5","tags":["web"]},{"url":"https://github.com/keras-team/keras/commit/d338a45204bdc787c8b3c4a9b82c1911cd52dedf","tags":["web"]},{"url":"https://github.com/keras-team/keras/releases/tag/v3.12.3","tags":["web"]},{"url":"https://github.com/keras-team/keras/releases/tag/v3.15.0","tags":["web"]},{"url":"https://github.com/advisories/GHSA-5gwj-m78q-7pq3","tags":["advisory"]},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/12xxx/CVE-2026-12481.json","tags":["advisory"]},{"url":"https://security-tracker.debian.org/tracker/CVE-2026-12481","tags":["advisory"]},{"url":"https://github.com/keras-team/keras","tags":["package"]},{"url":"https://pypi.org/project/keras","tags":["package"]},{"url":"https://ubuntu.com/security/CVE-2026-12481","tags":["report"]},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-12481","tags":["report"]}],"sources":{"cve":[{"id":"CVE-2026-12481","title":"Deserialization of Untrusted Data in keras-team/keras","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T20:36:05.003Z","modified":"2026-07-06T15:15:46.244Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high","assessor":"cna"}],"cwes":["CWE-502"],"updated_at":"2026-11-02T14:19:48.461833Z"}],"euvd":[{"id":"EUVD-2026-41600","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T20:36:05Z","modified":"2026-07-06T15:15:46Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high"}],"updated_at":"2026-11-02T18:20:13.172179Z"}],"ghsa":[{"id":"GHSA-5gwj-m78q-7pq3","title":"Keras: Lambda deserialization can bypass safe mode and execute code","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T21:31:36Z","modified":"2026-08-07T20:30:43Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high"}],"cwes":["CWE-502"],"updated_at":"2026-11-02T14:40:58.888637Z"}],"nvd":[{"id":"CVE-2026-12481","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T21:16:54.737Z","modified":"2026-10-25T15:10:00.147Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical","assessor":"primary"},{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high","assessor":"secondary"}],"cwes":["CWE-502"],"updated_at":"2026-11-02T14:12:27.541515Z"}],"osv":[{"id":"CVE-2026-12481","title":"Deserialization of Untrusted Data in keras-team/keras","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T20:36:05.003Z","modified":"2026-10-28T02:30:59.740924675Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high"}],"cwes":["CWE-502"],"updated_at":"2026-11-02T15:44:57.354194Z"},{"id":"DEBIAN-CVE-2026-12481","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T21:16:54.737Z","modified":"2026-08-27T23:05:49.890938929Z","cvss":[{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical"}],"updated_at":"2026-11-02T15:45:51.374409Z"},{"id":"GHSA-5gwj-m78q-7pq3","title":"Keras: Lambda deserialization can bypass safe mode and execute code","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T21:31:36Z","modified":"2026-08-10T11:40:41.687024935Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high"}],"cwes":["CWE-502"],"updated_at":"2026-11-02T15:46:38.553953Z"},{"id":"PYSEC-2026-3631","title":"Keras: Lambda deserialization can bypass safe mode and execute code","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-08-10T10:43:49.135174Z","modified":"2026-08-10T11:30:11.789165677Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high"}],"updated_at":"2026-11-02T15:52:56.847083Z"},{"id":"UBUNTU-CVE-2026-12481","description":"A vulnerability in keras-team/keras version 3.14.0 allows for arbitrary code execution due to improper handling of deserialization in the `Lambda` layer. Specifically, the `_raise_for_lambda_deserialization()` function fails to enforce the safe-mode guard when `safe_mode` is set to `None`, which is the default value when `from_config()` is called outside of a `SafeModeScope` context. This logic error conflates `None` (unset/default-deny) with `False` (explicitly disabled), bypassing the guard and allowing attacker-controlled `marshal` bytecode to be deserialized. Affected call sites include `keras.layers.deserialize(config)`, `keras.models.clone_model(model)`, and any direct invocation of `Lambda.from_config(config)` without an enclosing `SafeModeScope(True)`. This vulnerability can be exploited to achieve arbitrary OS-level code execution in the context of the server or user process.","status":"published","published":"2026-07-03T21:16:00Z","modified":"2026-07-09T21:30:36.762644175Z","cvss":[{"version":"3.0","score":8.8,"vector":"CVSS:3.0/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H","level":"high"},{"version":"3.1","score":9.8,"vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","level":"critical"}],"updated_at":"2026-11-02T16:56:00.592716Z"},{"id":"CLSA-2026-1790928189","title":"TuxCare security update for keras (4 CVEs)","description":"TuxCare rebuilt keras to address 4 CVEs. Fixed in 2.15.0.post3+tuxcare.","status":"published","published":"2026-10-02T08:03:09Z","modified":"2026-10-27T13:43:13.403546741Z","updated_at":"2026-11-02T15:43:48.65893Z"}]}}