Hugging Face Datasets before 5.2.0 contains an insecure deserialization vulnerability that allows attackers to execute arbitrary Python code by crafting a malicious pickle file loaded through the deprecated pandas packaged builder. Attackers can supply a pickle-backed dataset that is processed by Pandas._generate_tables via pandas.read_pickle() without validation, so when a victim calls datasets.load_dataset("pandas", data_files=...), the embedded reduce payload executes code with the victim process's privileges.
{
"cna_assigner": "VulnCheck",
"cwe_ids": [
"CWE-502"
],
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/57xxx/CVE-2026-57845.json"
}{
"extracted_events": [
{
"range": [
{
"introduced": "0"
},
{
"last_affected": "5.1.0"
}
],
"source": "AFFECTED_FIELD",
"strategy": "StandardRange"
},
{
"original_tag": "5.2.0",
"range": [
{
"fixed": "e45c5b38100d0a39f7d29e31431075051b97485b"
}
],
"source": "REFERENCES_TAG"
}
]
}