🔐 OpenAI представила GPT-6 Cyber — нейросеть, которая нашла 81 неизвестную уязвимость в Firefox Новая модель специально разработана для кибербезопасности. Она умеет искать ошибки в коде, проверять уязвимости и помогать специалистам устранять угрозы. Результаты уже впечатляют: • Обнаружила 81 ранее неизвестную серьёзную уязвимость в Firefox. • Помогла создать эксплойты для 948 уязвимостей ядра Linux. • Набрала 48,5% в ExploitGym против 32,1% у предыдущей GPT-5.6 Cyber. По словам OpenAI, в большинстве задач по кибербезопасности новинка превосходит даже GPT-6 Astra. Пока доступ к GPT-6 Cyber получит только ограниченный круг специалистов через API и Codex в рамках программы Daybreak Red. Источник
From the original source:
Telegram — GPT Time | Нейросети | ИИ
Read full article at Telegram — GPT Time | Нейросети | ИИOpenAI unveils GPT-6 Cyber — a neural network that discovered 81 unknown vulnerabilities in Firefox
🔐 OpenAI unveils GPT-6 Cyber — a neural network that discovered 81 unknown vulnerabilities in Firefox. The new model is specifically designed for cybersecurity. It can search for code errors, test vulnerabilities, and help specialists eliminate threats. The results are already impressive: •…
Telegram — GPT Time | Нейросети | ИИBy: GPT Time | Нейросети | ИИNovember 6, 2026 (1 hour ago)
Available in 4 languagesAI Summary & Key Takeaways
- 🔐 OpenAI unveils GPT-6 Cyber — a neural network that discovered 81 unknown vulnerabilities in Firefox
- The new model is specifically designed for cybersecurity
- It can search for code errors, test vulnerabilities, and help specialists eliminate threats
Research
KZ·AI 9A25Инсайт дня
OpenAI unveils GPT-6 Cyber — a neural network that discovered 81 unknown vulnerabilities in Firefox
Full text available in Russian
Tags:#OpenAI
Related stories
Hundreds of Artex AI hacking tools found publicly exposed to the web
Security researchers found hundreds of servers hosting the penetration testing tool Artex exposed to the internet, allowing anyone access to the hacking tool.
TechCrunch — AI
Meta introduces Spaces, a new way to collaborate with your Muse agent
Meta's Muse agents continue to get new features and integrations. Spaces promises to be a new way to interact with the AI helpers.
TechCrunch — AI
Free Chinese AI Models Are Getting Scary Good at Hacking
An experiment in which AI models were asked to break into simulated networks showed the hacking skills of DeepSeek and GLM to be not far behind OpenAI’s most advanced cyber model, which isn’t public.
Wired — Artificial Intelligence
AI News: New ChatGPT & Google Models (14 Updates Worth Knowing)
See what Optimizely’s Virtual Teammates can automate for you at http://optimizely.com/mattwolfe This week’s AI news covers OpenAI’s GPT-Image 2.6, Google Gemini Omni 1.2 Flash, and the latest AI tools for creating images, generating videos, and getting things done. Watch demos…
YouTube — Matt Wolfe