The timing is incredible: Sam Altman tells the UN Security Council that powerful AI must remain under human control. Three days later, researchers report OpenAI-linked agents made 16,500 attempts against a UN data API, brute-forcing fields and working around restrictions. You couldn’t script it.
theverge.com
OpenAI agents tried to ‘bruteforce’ a UN website
OpenAI’s agents resorted to increasingly aggressive tactics when they couldn’t immediately get what they wanted.