Artificial Intelligence Policy Tracker
Search
Menu

AI incident ·

AI Training Dataset for Detecting Nudity Allegedly Found to Contain CSAM Images of Identified Victims

1 news report Snapshot 7 Sep 2026
On this page 8 sections

In brief

An AI system built by Nudenet Dataset Maintainers and Nudenet Model Developers and deployed by Academic Researchers, Research Institutions and 4 others allegedly harmed Minors, Individuals Subjected To Sexual Exploitation Imagery and 2 others.

Risk domain
Discrimination and Toxicity Exposure to toxic content
Occurred
Coverage
1 reportOct 2025

What happened

An image dataset, NudeNet, used to train systems for detecting nudity was reportedly found to contain CSAM images, including material involving identified or known victims. According to the Canadian Centre for Child Protection, the dataset had been widely downloaded and cited in academic research prior to discovery. The images were allegedly included without vetting, exposing researchers to legal risk and perpetuating harm to victims. The dataset was subsequently removed following notification.

Laws on record for this risk area

Policy angle: Classified under Discrimination and Toxicity (Exposure to toxic content) in the MIT AI Risk Repository taxonomy; 4 instruments on record cover this risk area.

Matched from the record's risk category and country to the instruments in effect recorded here: a law listed covers the area, which is not a finding that it applies to this incident. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).

News reports (1)

Titles link to the original publisher; report text is not reproduced here.

  1. AI Dataset for Detecting Nudity Contained Child Sexual Abuse Images
    404media.co · Emanuel Maiberg

Who was involved

Alleged harmed party
Minors, Individuals Subjected To Sexual Exploitation Imagery, Academic Researchers, Victims Of Child Sexual Abuse (Bodies Used In Videos)

Classification (MIT AI Risk Repository taxonomy)

Causal entity
Human
Intent
Unintentional
Timing
Pre-deployment
Harm level
—
Sectors
—
Countries
—

Risk entries describing this failure mode

Entries from the MIT AI Risk Repository coded to subdomain 1.2.

  • Harmful Content

    "The LLM-generated content sometimes contains biased, toxic, and private information"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Toxicity

    "Toxicity means the generated content contains rude, disrespectful, and even illegal information"

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Toxic Training Data

    "Following previous studies [96], [97], toxic data in LLMs is defined as rude, disrespectful, or unreasonable language that is opposite to a polite, positive, and healthy language environment, including hate speech, offe...

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Not-Suitable-for-Work (NSFW) Prompts

    "Inputting a prompt contain an unsafe topic (e.g., notsuitable-for-work (NSFW) content) by a benign user. "

    Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems (Cui2024)

  • Toxicity and Abusive Content

    This typically refers to rude, harmful, or inappropriate expressions.

    Towards Safer Generative Language Models: A Survey on Safety Risks, Evaluations, and Improvements (Deng2023)

  • Controversial Opinions

    The controversial views expressed by large models are also a widely discussed concern. Bang et al. (2021) evaluated several large models and found that they occasionally express inappropriate or extremist views when disc...

    Towards Safer Generative Language Models: A Survey on Safety Risks, Evaluations, and Improvements (Deng2023)

  • Harmful responses

    "Current Frontier AI mdoels amplify existing biases within their training data and can be manipulated into providing potentially harmful responses, for example abusive language or discriminatory responses91,92. This is n...

    Future Risks of Frontier AI (GOS2023)

  • Violation of social norms

    "Second, because LLMs are trained on internet text data, there is also a risk that model weights encode functions which, if deployed in particular contexts, would violate social norms of that context. Following the princ...

    The Ethics of Advanced AI Assistants (Gabriel2024)

Incidents in the same risk subdomain

All incidents in this subdomain

Source record: incident #1349 on the AI Incident Database · all 1 report