🤖 OpenAI warns: future AI models may pose a serious cyber threat

⚠️ The company states that models are developing faster than anticipated, and the next generations already potentially carry a "high level of risk"

📉 In a special hacker test, GPT-5 was able to complete 27% of tasks, while the improved GPT-5.1 — already 76%, which directly indicates the increase in its capabilities in system breaches

🧊 Now every new model is considered high-risk by default

⚠️ OpenAI emphasizes: this is not a problem of one company — the entire industry is moving towards more powerful and potentially more dangerous AIs

📉 Cybersecurity risks are increasing along with the speed of technological development

🟣🟢🟠

#BNB #BTC #ETH