{"id":"EM-20261103-J9I2","url":"https://cyber.eventmonitor.net/incident/EM-20261103-J9I2/","type":"cve","title":"Stanza Vulnerability (CVE-2026-115878)","subtitle":"CVE-2026-115878","subject":"CVE-2026-115878","status":"Developing","severity":"high","affected":[],"headline":null,"born":"2026-11-03T21:17:02.377Z","firstSignal":"2026-11-03T20:29:16.000Z","lastSignal":"2026-11-03T21:17:02.377Z","updated":"2026-11-04T14:43:58.877Z","revision":1,"life":[{"status":"Born","at":"2026-11-03T21:17:02.377Z","why":"second independent source: C6 NVD CVE API 2.0"},{"status":"Developing","at":"2026-11-03T21:17:02.377Z","why":"signals keep joining"}],"changeLog":[{"at":"2026-11-03T20:29:16.000Z","what":"First signal: C26 GitHub Advisory Database"},{"at":"2026-11-03T21:17:02.377Z","what":"Born: second source, C6 NVD CVE API 2.0. Page EM-20261103-J9I2 created."}],"stages":[{"name":"Disclosed","at":"2026-11-03T21:17:02.377Z","detail":"C6 NVD published"},{"name":"Scored","at":"2026-11-03T21:17:02.377Z","detail":"CVSS 7.8 · EPSS 0.6%"},{"name":"Exploited","at":null,"detail":"No exploitation reported"},{"name":"Advisories","at":null,"detail":"None yet"},{"name":"Vendor Response","at":null,"detail":"None in sources yet"},{"name":"Resolved","at":null,"detail":"After 14 quiet days"}],"signals":[{"at":"2026-11-03T20:29:16.000Z","source":"C26","sourceName":"GitHub Advisory Database","role":"Database","severity":3,"title":"GHSA-gh9r-c94j-cmp5 / CVE-2026-115878: Stanza: RCE via pickle deserialization in the lemmatizer","text":"pip:stanza","url":"https://github.com/advisories/GHSA-gh9r-c94j-cmp5"},{"at":"2026-11-03T21:17:02.377Z","source":"C6","sourceName":"NVD CVE API 2.0","role":"Database","severity":3,"title":"CVE-2026-115878 (CVSS 7.8)","text":"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-i","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-115878"}],"crowd":[],"reference":{"nvd":{"published":"2026-11-03T21:17:02.377","score":7.8,"severity":"HIGH","desc":"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","product":"","seen":1793803438877},"kev":null,"epss":{"epss":0.00597,"pct":0.48705,"at":1793803438877},"p0":false},"mergedInto":null,"publisher":"Event Monitor Cyber"}