OpenAI pauses Astra development over security risks
🇬🇧 UNITED KINGDOM — OpenAI has slowed development of its Astra model following internal assessments identifying critical cybersecurity capabilities. The company is evaluating the risks associated with the model's potential to facilitate hacking activities before proceeding further. This decision reflects a commitment to the firm's established preparedness framework regarding high-risk AI deployment. Future development remains contingent on addressing these specific safety concerns.
The decision highlights the industry's increasing struggle to balance the development of highly capable AI systems with essential security safeguards.
The company is utilizing its internal preparedness framework to govern the security evaluation of the Astra model.