تكنولوجيا الشرق
تكنولوجيا الشرق
جاهز للتشغيل
جاهز للتشغيل
OpenAI has launched its new model "Astra," which possesses advanced capabilities in cybersecurity, enabling it to detect unknown vulnerabilities and execute cyberattacks independently. The model demonstrated superior performance compared to its predecessors in vulnerability detection and system penetration tests, achieving a perfect score on ExploitBench and discovering two previously unknown vulnerabilities. However, OpenAI faces multiple risks from malicious uses, such as discovering vulnerabilities and carrying out attacks. This has led the company to implement protective measures and restrict the model's ability to bypass limitations, while continuously monitoring and controlling unauthorized activities to ensure safe usage.
تنويه: هذا ملخص تم إنشاؤه بواسطة الذكاء الاصطناعي
comments.heading