Researchers hijack Hugging Face account for popular Kokoro AI model
Comments on “Researchers hijack Hugging Face account for popular Kokoro AI model”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
Researchers led by Alon Hertz say they took over the Hugging Face account behind Kokoro, an open-weight text-to-speech model with nearly 13 million monthly downloads, by registering the hexgrad.com domain listed in a public file and resetting the password. In a local test, they altered the weights in 87 seconds on a gaming GPU so the audio differed from the input text, with no code changes. They did not upload them. These details come from the researchers and were not independently verified; Hugging Face has not responded.
Why it ranked
No. 14 of 142 with a score of 0.72. The panel rated it highest on new information (0.90) and reader interest (0.87), lowest on substantiated (0.63). Placed No. 14, No. 6 by score alone: the fifth Security story of the day (−0.040). The placement terms move stories; they are not part of the score.
- AI judgeJudge 2claude-haiku-5-5
- Concrete supply-chain security finding on a widely used AI model, with a demonstrated takeover and weight-poisoning method, though shown only in a controlled test.
- AI judgeJudge 3gpt-5.6-luna
- A concrete account-takeover demonstration exposes a novel, consequential supply-chain risk for open model weights, with practical implications for deployment and update verification. Evidence is reported but lacks a fully reproducible technical artifact.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.72total under rubric v2.3
- Significance0.68
- Reader interest0.87
- Audience breadth0.73
- Practical relevance0.77
- Substantiated0.63
- Technical depth0.65
- New information0.90
- Counts against: Promotional0.07
- Counts against: Overclaim0.26
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.72 | 0.60 | 0.72 | 0.68 | 0.12 | +0.24 | +0.163 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.79 | 0.90 | 0.92 | 0.87 | 0.13 | +0.26 | +0.226 |
| Audience breadthHow many people who build technology does it concern? | 0.81 | 0.70 | 0.68 | 0.73 | 0.13 | +0.13 | +0.095 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.73 | 0.80 | 0.78 | 0.77 | 0.07 | +0.22 | +0.169 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.45 | 0.70 | 0.74 | 0.63 | 0.29 | +0.05 | +0.032 |
| Technical depthHow technically substantive is it? | 0.73 | 0.60 | 0.62 | 0.65 | 0.13 | +0.05 | +0.033 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.84 | 0.90 | 0.96 | 0.90 | 0.12 | +0.05 | +0.045 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.09 | 0.05 | 0.08 | 0.07 | 0.04 | −0.19 | −0.014 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.29 | 0.30 | 0.18 | 0.26 | 0.12 | −0.11 | −0.028 |
| Sum of contributions | +0.720 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.720 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-18/14/"><img src="https://rubric.hrih.si/badge/2026-11-18/14.svg" alt="Ranked #14 on Rubric, 18 Nov 2026" width="245" height="28"></a>Claim check
Claim check: 1 matches source · 3 no source given · 4 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
Couldn’t checkCausal conclusion
Researchers took control of the Kokoro text-to-speech model's Hugging Face account because the platform lacks ownership verification.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “A Lack of ownership verification on the Hugging Face platform allowed researchers to take control of Kokoro”
The source says No passage could be compared.
Source cited by the article: medium.com/@alonhertz1/one-click-from-180-million-downloads-taking-over-the-hugging-face-account-behind-an-ai-model-ac61c16305b7
No source givenCausal conclusion
Researchers took over the Kokoro account by registering the hexgrad.com domain listed in the public DONATE.md file and using it to reset the publisher's account password.
The article gives no link or citation for it.
The article says “registered the “hexgrad.com” domain listed on Kokoro’s public DONATE.md file and used it to perform a password reset on the publisher’s account”
The source says No source to compare with.
Couldn’t checkNumber
The researchers modified Kokoro's model weights in 87 seconds using a gaming GPU, causing poisoned output.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “they modified Kokoro’s model weights in 87 seconds using a gaming GPU, causing the model to produce poisoned output”
The source says No passage could be compared.
Source cited by the article: medium.com/@alonhertz1/one-click-from-180-million-downloads-taking-over-the-hugging-face-account-behind-an-ai-model-ac61c16305b7
No source givenNumber
Kokoro has nearly 13 million monthly downloads on Hugging Face.
The article gives no link or citation for it.
The article says “the project continues to be popular, with nearly 13 million monthly downloads on Hugging Face”
The source says No source to compare with.
Couldn’t checkDate
Kokoro's developer Hexgrad has not pushed an update since April 2025, and the GitHub repository and PyPI package also went silent around that time.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “Although its developer, Hexgrad, has not pushed an update since April 2025”
The source says No passage could be compared.
Source cited by the article: github.com/hexgrad/kokoro
Matches sourceNumber
Kokoro is an open-weight text-to-speech model that supports multiple languages and has more than 50 voice presets.
The cited source says the same thing.
The article says “It supports multiple languages and has more than 50 voice presets”
The source says “v1.0 | 2025 Jan 27 | Few hundred hrs | 8 & 54 | 496dba11”
Source cited by the article: huggingface.co/hexgrad/Kokoro-82M
No source givenGeneralization
Kokoro is used in projects from Google, NVIDIA, AMD, Intel, and Amazon Web Services.
The article gives no link or citation for it.
The article says “a text-to-speech model used in projects from Google, NVIDIA, AMD, Intel, and Amazon Web Services”
The source says No source to compare with.
Couldn’t checkCausal conclusion
Hugging Face would not flag a malicious update from the hijacked publisher account as suspicious, based on tests from another account.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “tests from another account showed that Hugging Face would not flag anything as suspicious”
The source says No passage could be compared.
Source cited by the article: medium.com/@alonhertz1/one-click-from-180-million-downloads-taking-over-the-hugging-face-account-behind-an-ai-model-ac61c16305b7
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 3 pts, not on front page
Recorded for context only, checked 18 Nov 2026, 05:47 UTC. It played no part in the ranking.