OpenAI launches classifier to detect AI-generated child abuse imagery
Original title: OpenAI wants to stop deepfakes from overwhelming child abuse tiplines
OpenAI says the tool flags 48 percent fewer images while doubling the share that contain abusive material, with free access for vetted groups.
At a Washington briefing, OpenAI announced a classifier it says can better detect abusive imagery of children, including AI-generated material, and offered free access to vetted organizations. NCMEC’s CyberTipline received 400,000 generative AI-related reports in 2025, up from 67,000 in 2024. ChatGPT users who try to generate CSAM will see a new warning, a 24-hour pause on a first offense, and a possible permanent account closure. OpenAI says it reports all CSAM instances to NCMEC.
Unconfirmed: The false-positive reduction, detection-rate claims, and “first broadly available” AI-generated flag are OpenAI’s own statements.
- CyberTipline generative AI reports rose from 67,000 in 2024 to 400,000 in 2025.
- OpenAI says identified images were twice as likely to contain abusive material.
- A first ChatGPT CSAM attempt brings a 24-hour pause and a warning.
- The Take It Down Act requires removal of nonconsensual intimate images within 48 hours.
- Eight states sued OpenAI last month over separate alleged harms to children.
- Vendors
- OpenAI
- Products
- ChatGPT
- Organizations
- NCMEC
- Countries
- United States
From the original article
3 min read at theverge.com
OpenAI is announcing a tool that it says will help child sexual abuse investigators cut through the noise to find harmful material they can act on quickly. At a Monday press briefing in Washington, DC, OpenAI announced a classifier it says can better detect abusive imagery of kids, including whether it's AI-generated.
By Lauren Feiner · theverge.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.