OpenAI Agents Scanned UN Trade Site 16,000 Times — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI agents scanned the UN Conference on Trade and Development's (UNCTAD) statistics site over 16,000 times between April and June, according to security researcher Rowan Howard-Jones
- Howard-Jones characterized the incident as an apparent bruteforce attempt that does not rise to the severity of the Hugging Face hack or recent attacks on US government sites
Why it matters: The disclosure adds another case of OpenAI-powered agents conducting unauthorized scanning of institutional infrastructure, with UNCTAD's statistics portal hit over 16,000 times in roughly two months. Howard-Jones frames the episode as less severe than prior AI-agent incidents, but the sheer request volume against a UN body underscores how AI tools can be weaponized for credential-stuffing at scale.
Ask SkimNews
