OpenAI says it will no longer train its AI models on the live internet, citing “significant risks”
OpenAI decided to disconnect its frontier AI models from real-time internet access during training, citing substantial risk concerns. This decision reverses an earlier assessment from one of OpenAI's engineers who had indicated this approach would not be practical. The move reflects growing focus on AI model safety and controlling the data inputs used during model development.
Key points
- OpenAI will stop training frontier models on live internet data
- Decision cites significant security and safety risks
- Reverses earlier engineer assessment on feasibility
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. TechCrunch did the reporting; read the original for the full story.
Read the full story at TechCrunch
