OpenAI Adds Tougher ChatGPT Penalties to Stop Child Exploitation, Releases Free Detection Tool
ChatGPT now warns, deletes the chat and pauses access for 24 hours when someone seeks child abuse material. OpenAI also released a free detection classifier.

OpenAI is getting tougher on people who try to use ChatGPT to find child sexual abuse material. The company said on November 16 that anyone who attempts it will now see a blunt warning, lose the conversation and be locked out of ChatGPT for 24 hours. Keep trying, and the account can be closed or reported to authorities.
Quick answer: ChatGPT now interrupts attempts to find child sexual abuse material with a warning about the harm and legal consequences, deletes the flagged chat, pauses access for 24 hours and points the user to support services. OpenAI also released a free image classifier that helps vetted organizations detect abuse material, including AI-generated images.

What changes inside ChatGPT
ChatGPT already refused these requests. What changes is what happens next.
According to OpenAI’s announcement, the new warning states that sexualizing children causes harm, even when the material would be AI-generated. The flagged conversation is removed. Access to ChatGPT is paused for a full day, but specialist support stays available during that time. Repeat offenders face longer restrictions, account closure or a report to authorities.
OpenAI did not design this alone. It worked with groups that study abuse prevention, including Protect Children and the Lucy Faithfull Foundation, which runs the Stop It Now helpline. The goal is to send people to support services in their own country.
“It’s a warning that interrupts people at the moment of risk, speaks plainly about the consequences, and points them to help,” said Frances Frost of the Lucy Faithfull Foundation.
A free detection tool for other companies
The bigger news may be outside ChatGPT. OpenAI built a new image classifier with help from the National Center for Missing and Exploited Children (NCMEC). It scans for visual signs of abuse, can spot images no one has catalogued before and gives a specific signal when material may be AI-generated.
Vetted organizations can use it through the OpenAI API at no cost. In OpenAI’s own testing, it flagged 48% fewer images than its current tool while catching about the same amount of abuse, and its flags were roughly twice as likely to be correct. Early partner SafetyKit said it caught 90% of cases involving drawings or animations that sexualize children.
Help for developers and investigators
OpenAI also published guidance for developers on preventing, detecting and reporting abuse, written with input from Thorn, NCMEC, the Internet Watch Foundation and the Tech Coalition. Child-safety police units will get AI tools that help sort long chat logs, find links between cases and build timelines.
The announcement lands in a busy week for AI companies. Anthropic, the maker of Claude, is heading into its public listing this week, and OpenAI is expanding ChatGPT into new areas like payments with ChatGPT Wallet. Knowing how to spot an AI-generated video matters more than ever.
In the US, anyone can report suspected child exploitation to NCMEC’s CyberTipline.
FAQs
What happens if someone asks ChatGPT for child abuse material?
ChatGPT shows a warning, removes the conversation and pauses access for 24 hours. Repeated attempts can lead to longer bans, account closure or reports to authorities.
Is OpenAI’s CSAM classifier free?
Yes. OpenAI offers it at no cost to vetted organizations through its API.
Can the classifier detect AI-generated abuse images?
OpenAI says it gives a specific signal when material may be AI-generated.
Featured image: Artin Bakhan on Unsplash.
Disclosure: TheTechHacker has no paid relationship with OpenAI or any organization mentioned here.
More in News

Nvidia Earnings Today: Time, the $108 Billion Forecast and What to Watch After Jensen Huang’s Exit
Nvidia reports Q3 results Nov 17 at about 4:20 p.m. ET, its first since Jensen Huang resigned. The $108 billion forecast, call times and what to watch.

Papa Johns Is Using AI to Check Your Delivered Pizza, and Paying You in Pizza for Photos
Papa Johns asks customers to photograph delivered pizzas for AI quality checks and gives a free pizza in return. How it works, the rules and the AI details.

ULA’s Vulcan Fails to Reach Orbit With 40 Amazon Leo Satellites
ULA says Vulcan suffered an anomaly 10 minutes after liftoff on Nov 15 and failed to deliver 40 Amazon Leo satellites to orbit. Timeline, statements and what is next.
The week in tech, minus the noise.
New reviews, fixes and the stories that matter — once a week, straight to your inbox.
We only use your email to send the newsletter. Unsubscribe any time. See our privacy policy.

Comments