Particle.news
Download on the App Store

OpenAI Reaffirms Youth Safety Policies as It Preps Human-Like ChatGPT and Halts MLK Jr. Videos in Sora

The company pairs product changes with pledges on mental health safeguards to address growing safety concerns.

Overview

  • CEO Sam Altman said the company has not relaxed any mental health policies and stated that minors require serious protection when using AI.
  • OpenAI plans to release a version of ChatGPT in the coming weeks that can adopt a more human-like style on request, including emojis and a friendly tone.
  • Starting in December, the company intends to allow more explicit content for adults who pass age verification, while teen accounts begin at 13 and include parental controls.
  • OpenAI said it is training its models to better recognize signs of user stress following complaints about safety in chatbot interactions.
  • OpenAI temporarily blocked Sora videos depicting Martin Luther King Jr. at the request of his foundation after reports that users generated disrespectful portrayals.