OpenAI Says Astra AI Can Develop Cyberattacks Without Human Help

OpenAI says its new Astra artificial-intelligence model has reached the company’s “Critical” cybersecurity threshold. According to information reported by CoinDesk, the designation means Astra can identify previously unknown vulnerabilities and create methods to exploit them across hardened systems. OpenAI also says the model can build attacks without human assistance, highlighting its ability to perform advanced cybersecurity-related tasks autonomously. The announcement concerns the model’s capabilities and cybersecurity classification; the supplied information does not indicate any effect on cryptocurrency prices, blockchain networks or the digital-asset market. Astra’s classification marks the first time an OpenAI model has reached the “Critical” level under the company’s cybersecurity framework.