OPENAI

The algorithm has joined the neighbourhood watch

OpenAI has announced new safety measures to tackle child sexual abuse material (CSAM), including AI-generated images.

The update introduces a new image detection tool, developed with the National Center for Missing and Exploited Children (NCMEC), that can identify signs of abuse in previously unseen images.

The company is also introducing three additional measures:

  • Users attempting to access content that sexualises children will receive warnings, support resources and a 24-hour account suspension. Repeated violations could lead to permanent bans.

  • Instructions to help developers prevent, detect and report abusive material.

  • AI-powered investigation tools, data analysis features and training to help authorities investigate child exploitation cases.

A better filter, a bigger rulebook and a shorter leash

In early tests, the tool flagged 48% fewer images for review while detecting the same number of abusive images.

OpenAI plans to offer it free to selected companies.

The announcement comes amid growing scrutiny of AI safety, including lawsuits against OpenAI and concerns about the impact of chatbots on younger users.

A use for AI that we can all get behind. -MV