OpenAI has decided to pause the internal activities surrounding its artificial intelligence model in development, named Astra. The decision was made after the company was unable to rule out critical cyber capabilities according to the safety standards outlined in its evaluation framework.
The suspension highlights the growing focus on stringent safety protocols and risk assessment for advanced artificial intelligence models, specifically regarding potential vulnerabilities or misuse related to critical cybersecurity capabilities during the development phase.
In-depth summary · AI, neutral