OpenAI’s Newest Model Triggers a ‘Critical’ Warning‘

OpenAI has delayed the release of its latest, most powerful AI model, Astra, after internal testing revealed the system may possess cybersecurity capabilities so advanced that it could be classified as “Critical,” the highest risk category under the company’s Preparedness Framework.

The company behind the ChatGPT chatbot announced Friday that it is restricting Astra’s access to the internet and outside tools, strengthening protections around the model, and pausing internal activities that do not meet heightened security requirements while additional evaluations are completed.

"We are sharing this because we believe it’s important to be transparent with the public and the safety and security communities about this potential shift in capabilities," OpenAI said in a statement posted on its website. "Accordingly, we have scaled up robustness testing of our safeguards and security controls so that they are appropriate for a deployment of these capabilities. Internally, we have also taken steps so that further development of this model happens safely and securely."

Please select this link to read the complete article from Inc.