What we know
OpenAI has introduced a new image classifier designed to detect child sexual abuse material (CSAM) and added warnings within ChatGPT to address attempts to access such content. ChatGPT will impose a 24-hour ban on accounts that try to access CSAM as a warning measure. OpenAI’s new image classifier targets CSAM detection. ChatGPT enforces a 24-hour temporary ban on accounts attempting to access CSAM content.
Why it matters
Detecting and preventing CSAM is a critical challenge for AI platforms, especially as generative models can be exploited to create or share illegal content. OpenAI’s new tools aim to reduce the risk of misuse by identifying harmful images and discouraging attempts to access CSAM through temporary account restrictions. These efforts reflect growing industry focus on safeguarding AI systems from facilitating abuse.
What is still unknown
Details about the classifier’s accuracy, the range of languages supported, and when these features will be broadly available remain unclear. No independent tests of effectiveness or user impact have been reported.

