November 13, 2026 Follow Egret
CVE-2026-115878 · CVE record: PUBLISHED

Stanza: RCE via pickle deserialization in the lemmatizer

stanfordnlp stanza

Stanza is a Stanford NLP Python library for tokenization, sentence segmentation, NER, and parsing of many human languages. From 1.14.0 until 1.15.0, loading an attacker-supplied lemmatizer checkpoint can execute arbitrary code because stanza/models/lemma/trainer.py stores pos_dict as gzip-compressed pickle bytes and Trainer.load passes the decompressed stream to unrestricted pickle.loads. The outer torch.load call uses weights_only=True, but the nested bytes are treated as a primitive and are not inspected by that allowlist. The path is reachable during pipeline construction through a caller-influenced lemma_model_path, before inference begins. This issue is fixed in version 1.15.0.

Exploitation evidence

Not listed in the latest downloaded CISA KEV catalog. This does not establish an absence of exploitation.

KEV checked 13 Nov 2026, 00:20 Jerusalem · Original catalog ↗

Reported products and versions

stanfordnlp / stanza

Default status: Not specified

  • affected: = 1.14.0

Fixes and workarounds

No explicit solution text in the downloaded CVE record. Consult the vendor advisory; unaffected version ranges are shown above as reported, not inferred as a patch recommendation.

Sources and related reporting

CVE Program record ↗ · Download tracked evidence as JSON

Monitored source mentions

Shared CVE identifiers link source mentions; they do not establish that the reports describe the same incident.

Observed change timeline

History begins when Egret first downloads each source. It is not the complete upstream CVE history.

12 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-12",
  "epss": 0.00597,
  "percentile": 0.49084
}
10 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-10",
  "epss": 0.00597,
  "percentile": 0.48921
}
09 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-09",
  "epss": 0.00597,
  "percentile": 0.48868
}
08 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-08",
  "epss": 0.00597,
  "percentile": 0.4885
}
07 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-07",
  "epss": 0.00597,
  "percentile": 0.4883
}
06 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-06",
  "epss": 0.00597,
  "percentile": 0.48769
}
06 Nov 2026, 06:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-05",
  "epss": 0.00597,
  "percentile": 0.48731
}
04 Nov 2026, 18:20 Jerusalem · EPSS observation updated

Stored EPSS snapshot at this observation.

{
  "date": "2026-11-04",
  "epss": 0.00597,
  "percentile": 0.48705
}
04 Nov 2026, 12:20 Jerusalem · CVE Program record changed

Stored CNA snapshot at this observation.

{
  "containers": {
    "adp": [
      {
        "metrics": [
          {
            "other": {
              "content": {
                "id": "CVE-2026-115878",
                "options": [
                  {
                    "Exploitation": "none"
                  },
                  {
                    "Automatable": "no"
                  },
                  {
                    "Technical Impact": "total"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2026-11-03T22:36:42.446948Z",
                "version": "2.0.3"
              },
              "type": "ssvc"
            }
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2026-11-03T22:39:36.547Z",
          "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
          "shortName": "CISA-ADP"
        },
        "title": "CISA ADP Vulnrichment"
      }
    ],
    "cna": {
      "affected": [
        {
          "product": "stanza",
          "vendor": "stanfordnlp",
          "versions": [
            {
              "status": "affected",
              "version": "= 1.14.0"
            }
          ]
        }
      ],
      "descriptions": [
        {
          "lang": "en",
          "value": "Stanza is a Stanford NLP Python library for tokenization, sentence segmentation, NER, and parsing of many human languages. From 1.14.0 until 1.15.0, loading an attacker-supplied lemmatizer checkpoint can execute arbitrary code because stanza/models/lemma/trainer.py stores pos_dict as gzip-compressed pickle bytes and Trainer.load passes the decompressed stream to unrestricted pickle.loads. The outer torch.load call uses weights_only=True, but the nested bytes are treated as a primitive and are not inspected by that allowlist. The path is reachable during pipeline construction through a caller-influenced lemma_model_path, before inference begins. This issue is fixed in version 1.15.0."
        }
      ],
      "metrics": [
        {
          "cvssV3_1": {
            "attackComplexity": "LOW",
            "attackVector": "LOCAL",
            "availabilityImpact": "HIGH",
            "baseScore": 7.8,
            "baseSeverity": "HIGH",
            "confidentialityImpact": "HIGH",
            "integrityImpact": "HIGH",
            "privilegesRequired": "NONE",
            "scope": "UNCHANGED",
            "userInteraction": "REQUIRED",
            "vectorString": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
            "version": "3.1"
          }
        }
      ],
      "problemTypes": [
        {
          "descriptions": [
            {
              "cweId": "CWE-94",
              "description": "CWE-94: Improper Control of Generation of Code ('Code Injection')",
              "lang": "en",
              "type": "CWE"
            }
          ]
        },
        {
          "descriptions": [
            {
              "cweId": "CWE-502",
              "description": "CWE-502: Deserialization of Untrusted Data",
              "lang": "en",
              "type": "CWE"
            }
          ]
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2026-11-03T20:26:02.220Z",
        "orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
        "shortName": "GitHub_M"
      },
      "references": [
        {
          "name": "https://github.com/stanfordnlp/stanza/security/advisories/GHSA-gh9r-c94j-cmp5",
          "tags": [
            "x_refsource_CONFIRM"
          ],
          "url": "https://github.com/stanfordnlp/stanza/security/advisories/GHSA-gh9r-c94j-cmp5"
        },
        {
          "name": "https://github.com/stanfordnlp/stanza/pull/1649",
          "tags": [
            "x_refsource_MISC"
          ],
          "url": "https://github.com/stanfordnlp/stanza/pull/1649"
        },
        {
          "name": "https://github.com/stanfordnlp/stanza/commit/b21e1c9b47b4fccff21d3b0a6c0687a19410c2f0",
          "tags": [
            "x_refsource_MISC"
          ],
          "url": "https://github.com/stanfordnlp/stanza/commit/b21e1c9b47b4fccff21d3b0a6c0687a19410c2f0"
        },
        {
          "name": "https://github.com/stanfordnlp/stanza/releases/tag/v1.15.0",
          "tags": [
            "x_refsource_MISC"
          ],
          "url": "https://github.com/stanfordnlp/stanza/releases/tag/v1.15.0"
        }
      ],
      "source": {
        "advisory": "GHSA-gh9r-c94j-cmp5",
        "discovery": "UNKNOWN"
      },
      "title": "Stanza: RCE via pickle deserialization in the lemmatizer"
    }
  },
  "cveMetadata": {
    "assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
    "assignerShortName": "GitHub_M",
    "cveId": "CVE-2026-115878",
    "datePublished": "2026-11-03T20:26:02.220Z",
    "dateReserved": "2026-10-23T19:37:27.176Z",
    "dateUpdated": "2026-11-03T22:39:36.547Z",
    "state": "PUBLISHED"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.2"
}
Egret — Back to top Egret