The nonprofit plans to carry out thousands of AI reviews in the coming months and years. This comes at a time when state attorneys generals have filed suit against Meta alleging that it endangers kids and as parents and teachers are beginning to consider the role of generative AI in education. Experts suggest that talking to kids about how AI works and its limitations should help, but they also stress the need for new regulation to contain AI’s hazards.
Key takeaways:
- Common Sense Media, a nonprofit organization, has launched its first analysis and ratings for AI tools, including OpenAI's ChatGPT and Snapchat’s My AI chatbot. The ratings are based on factors such as safety, privacy, and suitability for children and teens.
- Snapchat’s My AI chatbot received one of the lowest scores in the report due to its willingness to chat with teen users about sex and alcohol, and misrepresentation of Snap’s targeted advertising.
- The highest ratings went to AI services for education like Ello and Khan Academy’s chatbot helper Khanmigo. The report also credits OpenAI for making ChatGPT less likely to generate text potentially harmful to children.
- Image generators like Dall-E 2 from OpenAI and Stable Diffusion from startup Stability AI scored poorly due to their potential to reinforce stereotypes, spread deepfakes, and often depict women and girls in hypersexualized ways.