OpenAI agents tried to ‘bruteforce’ a UN website

OpenAI agents were used to scan the UNCTAD website over 16,000 times between April and June, highlighting concerns about the potential misuse of AI for brute-force attacks.