OpenAI has announced that its next AI model, Astra, is the first to hit the company’s “critical cybersecurity capability” benchmark, meaning it can autonomously discover and exploit hidden vulnerabilities in highly secure systems. The firm noted that because Astra can operate without human prompting, it poses a higher risk than prior models and will therefore be subject to stricter safeguards before it is released. OpenAI’s blog post explains that Astra’s advanced probing abilities could be used for both defensive testing and malicious exploitation, prompting the company to emphasize responsible deployment and enhanced security protocols. This development raises questions about how AI tools that can automatically uncover security weaknesses will be regulated and monitored in the future.
OpenAI's forthcoming artificial intelligence model Astra is the company's first to meet its "critical cybersecurity capability" threshold and will require stronger safeguards before release, the AI firm said Tuesday. In a blog post, OpenAI said Astra can find and exploit previously unknown security flaws across "many well-protected systems" without a human prompt to do so....
