OpenAI partially halts development of AI model Astra

OpenAI has partially halted the development of its new AI model Astra after security concerns arose. Reports indicate that the model demonstrates critical capabilities that could potentially lead to cyberattacks on highly secure targets.

In response to these findings, OpenAI announced that it would tighten controls for new AI systems. The company intends to monitor internal processes more rigorously to identify potential risks early. A timeline for resuming testing is currently unknown.