Denial of Service via HDF5 Shape Bomb in keras.models.load_model() in keras-team/keras
A vulnerability in keras-team/keras versions <= 3.15.0 allows for a denial of service (DoS) attack when loading malicious .keras model files via the keras.models.load_model() function. The H5IOStore.__getitem__ method in keras/src/saving/saving_lib.py does not validate the shape or size of datasets, leading to unbounded memory allocation. A specially crafted .keras file can exploit this flaw to trigger an out-of-memory (OOM) condition, causing the process to be terminated (exit code 137). This issue bypasses the fix for CVE-2026-0897, which only addressed a similar vulnerability in KerasFileEditor. The attack vector includes poisoned models from public repositories or malicious model registries, posing a risk to machine learning pipelines that process untrusted models.
- Products
- Not specified
- Affected versions
- Not specified
- Fixed versions
- Not specified
- CVEs
- CVE-2026-12570
- Confirmed fixed CVEs
- Not specified
- Updated in this feed
- 4 Nov 2026, 11:19 UTC · Revision 1
Source evidence
Denial of Service via HDF5 Shape Bomb in keras.models.load_model() in keras-team/keras
core-osv · Original sourcePublished 10 Aug 2026, 06:29 UTC
A vulnerability in keras-team/keras versions <= 3.15.0 allows for a denial of service (DoS) attack when loading malicious .keras model files via the keras.models.load_model() function. The H5IOStore.__getitem__ method in keras/src/saving/saving_lib.py does not validate the shape or size of datasets, leading to unbounded memory allocation. A specially crafted .keras file can exploit this flaw to trigger an out-of-memory (OOM) condition, causing the process to be terminated (exit code 137). This issue bypasses the fix for CVE-2026-0897, which only addressed a similar vulnerability in KerasFileEditor. The attack vector includes poisoned models from public repositories or malicious model registries, posing a risk to machine learning pipelines that process untrusted models.
osv:CVE-2026-12570
Complete source record
{
"id": "e0fa8c288e6c72731aaf5fc75435edff4e47f0c0ad424425614eaafc833229d6",
"url": "https://osv.dev/vulnerability/CVE-2026-12570",
"method": "structured",
"payload": {
"kind": "advisory",
"title": "Denial of Service via HDF5 Shape Bomb in keras.models.load_model() in keras-team/keras",
"summary": "A vulnerability in keras-team/keras versions <= 3.15.0 allows for a denial of service (DoS) attack when loading malicious .keras model files via the keras.models.load_model() function. The H5IOStore.__getitem__ method in keras/src/saving/saving_lib.py does not validate the shape or size of datasets, leading to unbounded memory allocation. A specially crafted .keras file can exploit this flaw to trigger an out-of-memory (OOM) condition, causing the process to be terminated (exit code 137). This issue bypasses the fix for CVE-2026-0897, which only addressed a similar vulnerability in KerasFileEditor. The attack vector includes poisoned models from public repositories or malicious model registries, posing a risk to machine learning pipelines that process untrusted models.",
"evidence": "osv:CVE-2026-12570",
"products": [],
"severity": "medium",
"structured": {
"id": "CVE-2026-12570",
"aliases": [
"CVE-2026-12570",
"GHSA-74m6-m3xx-3vmj",
"PYSEC-2026-3856"
],
"metrics": [
{
"type": "CVSS_V3",
"score": "CVSS:3.0/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H"
}
],
"affected": [
{
"ranges": [
{
"repo": "https://github.com/keras-team/keras",
"type": "GIT",
"events": [
{
"introduced": "0"
},
{
"fixed": "4933ea4a5b3fcc24ceacdc276f5bb5dfbd06756c"
}
],
"database_specific": {
"extracted_events": [
{
"range": [
{
"introduced": "0"
},
{
"last_affected": "3.15.0"
}
],
"source": "DESCRIPTION"
},
{
"range": [
{
"fixed": "4933ea4a5b3fcc24ceacdc276f5bb5dfbd06756c"
}
],
"source": "REFERENCES_COMMIT"
}
]
}
}
],
"versions": [
"v3.14.0",
"v3.12.0",
"v3.11.0",
"v3.10.0",
"v3.9.0",
"v3.8.0",
"v3.7.0",
"v3.6.0",
"v3.5.0",
"v3.4.1",
"v3.4.0",
"v3.3.3",
"v3.3.2",
"v3.3.1",
"v3.3.0",
"v3.2.1",
"v3.2.0",
"v3.1.1",
"v3.1.0",
"v3.0.5",
"v3.0.4",
"v3.0.3",
"v3.0.2",
"v3.0.1",
"v3.0.0",
"v2.9.0-rc0",
"v2.8.0-rc0",
"v2.7.0-rc0",
"v2.6.0-rc0",
"2.4.0",
"2.3.1",
"2.3.0",
"2.2.4",
"2.2.3",
"2.2.2",
"2.2.1",
"2.2.0",
"2.1.6",
"2.1.5",
"2.1.4",
"2.1.3",
"2.1.2",
"2.1.1",
"2.1.0",
"2.0.9",
"2.0.8",
"2.0.7",
"2.0.6",
"2.0.5",
"2.0.4",
"2.0.3",
"2.0.2",
"2.0.1",
"2.0.0",
"1.2.0",
"1.1.1",
"1.1.0",
"1.0.8",
"1.0.7",
"1.0.6",
"1.0.5",
"1.0.4",
"1.0.3",
"1.0.2",
"1.0.1",
"1.0.0",
"0.3.2",
"0.3.1",
"0.3.0",
"0.2.0",
"0.1.3",
"0.1.2",
"0.1.1",
"0.1.0"
],
"database_specific": {
"source": "https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-12570.json"
}
}
],
"modified": "2026-10-28T02:30:34.437908232Z",
"provider": "osv",
"references": [
"https://huntr.com/bounties/a064f475-780a-409a-82f7-678512f27ad8",
"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/12xxx/CVE-2026-12570.json",
"https://nvd.nist.gov/vuln/detail/CVE-2026-12570",
"https://github.com/keras-team/keras/commit/4933ea4a5b3fcc24ceacdc276f5bb5dfbd06756c"
],
"provider_metadata": {
"cwe_ids": [
"CWE-770"
],
"cna_assigner": "@huntr_ai",
"unresolved_ranges": [
{
"extracted_events": [
{
"range": [
{
"fixed": "3.12.3, 3.15.0"
}
],
"source": "AFFECTED_FIELD",
"strategy": "SplitRange"
}
]
}
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/12xxx/CVE-2026-12570.json"
}
},
"related_cves": [
{
"value": "CVE-2026-12570",
"evidence": "/id"
}
],
"security_relevant": true,
"severity_evidence": "/severity/0/score"
},
"source_id": "core-osv",
"document_id": 501985,
"published_at": "2026-08-10T06:29:56.111+00:00"
}