OpenAI has announced that it has suspended development of certain elements of its new Astra model. An internal audit revealed a significant leap in its coding and cybersecurity capabilities — details that were powerful enough to raise concerns about the development’s security.
OpenAI has announced that it has suspended development of certain elements of its new Astra model. An internal audit revealed a significant leap in its coding and cybersecurity capabilities — details that were powerful enough to raise concerns about the development’s security.
OpenAI said that Astra, which is still in development, has reached a “critical cybersecurity threshold.” This means it is capable of independently detecting and executing cyberattacks on real-world systems that are usually well-protected. This has led to the activation of additional security measures, TechCrunch reports .
“While we continue to test and evaluate this model, our preliminary tests indicate such high performance that we cannot currently rule out the possibility of reaching the Critical Capability Level,” OpenAI said. “Astra is our future model and has not been used to exploit the Hugging Face vulnerabilities.”
OpenAI said it was sharing this information because it believed it was “important to be transparent with the public and security communities about this potential leap in capabilities.”
The company also said it is taking steps, including implementing stricter security controls and suspending internal Astra-related activities that do not comply with these enhanced restrictions. OpenAI added that it is working with relevant government agencies and “selected AI security organizations” to test the model’s capabilities.
