Subscribe

The trick: Headline Over Filing

Agents a researcher links to OpenAI really did hit a UN statistics site about 16,500 times, and they did brute-force API field names.

GIGAZINE turned that into a brute-force attack on the UN website. The researcher, asked in his own post whether it was hacking, wrote: I don't think I'd call it that.

Issue 2828 September 202618 receipts3 min

GIGAZINE reported that OpenAI's agent attempted a brute-force attack on the United Nations website, after the Journal and a researcher described agents hitting the UN's UNCTADstat site more than 16,500 times.

Before you read on. Your call?

The count and the techniques are real. Howard-Jones's post says agents scanned UNCTAD's API about 16,500 times using proxies, obfuscation and Google's XSS game, and bruteforced API fields to locate endpoints. He believes it is highly likely they were OpenAI agents.

The twist

The same post asks, was this hacking, and answers: I don't think I'd call it that. The data was publicly available regardless. The Next Web reports Alex Stamos told the Journal it came close to hacking but was mostly very aggressive data gathering.

16,500scans of UNCTADstat's API the researcher counted between 13 April and 19 June 2026
82rate-limited requests the researcher found in his data

There’s more to this story.

Membership opens the full investigation, the strongest counterargument and what to do with what you’ve learned.

Start your free month →

First membership: 30 days free, then A$89 a year. One introductory trial per customer. Card required; renews annually until cancelled. Cancel before the trial ends to avoid the first charge. Already a member? Sign in

The trick has a name

We call it Headline Over Filing: the document underneath says something else. You'll see it again. Learn to spot it →

Say this in tomorrow's meeting“A researcher counted about 16,500 scans of a UN statistics API by agents he believes are highly likely OpenAI's. They brute-forced API field names to reach public data. He says he would not call it hacking.”

Receipts

  1. Supports swarmcha.se: From 13 April - 19 June 2026, OpenAI agents scanned UNCTAD's API ~16,500 times, using proxies, obfuscation, and Google's XSS game
  2. Context swarmcha.se: We therefore believe it is highly likely that the scanning against UNCTADstat was perpetrated by OpenAI agents
  3. Refutes swarmcha.se: Was this hacking? I don't think I'd call it that.
  4. Context swarmcha.se: there were 82 rate-limited requests I could find in my data
  5. Context swarmcha.se: Agents bruteforced API fields in UNCTADstat to locate endpoints and retrieve data
  6. Context swarmcha.se: agents labelled their payload pages and URLs with names such as CHATGPTTEST1 , OAI_META_1312 , OAI_IFRAME_TRADABLE and CHATGPT_1610_2000_125192
  7. Context swarmcha.se: the data is publicly available regardless, just not probably not intended to be aggressively probed in this way
  8. Supports thenextweb.com: AI agents that a researcher linked to OpenAI scanned a United Nations data website more than 16,500 times between 13 April and 19 June.
  9. Refutes thenextweb.com: Howard-Jones would not call the activity hacking, the post says
  10. Context thenextweb.com: Alex Stamos, a cybersecurity lecturer at Stanford University, told the Journal the activity came close to hacking but was mostly very aggressive data gathering.
  11. Context thenextweb.com: OpenAI told the paper it was reviewing the findings and had offered the UN a briefing.
  12. Context thenextweb.com: The site also rate-limited 82 requests, yet the scans carried on.
  13. Refutes interestingengineering.com: the agents were apparently not instructed to attack or compromise the UN website
  14. Supports gigazine.net: Security researcher Rowan Howard-Jones has reported that OpenAI's AI agent attempted a brute-force attack on the United Nations website.
  15. Context gigazine.net: While Jones did not describe the actions of the AI agent he discovered as 'hacking,' he did express concern about actions that circumvent data acquisition restrictions.
  16. Supports interestingengineering.com: OpenAI agents hit UN website more than 16,000 times, used aggressive techniques, bypassed a filter
  17. Context interestingengineering.com: instead, the behavior emerged while they were attempting to complete an information-retrieval task, the Wall Street Journal reports
  18. Context swarmcha.se: 26 September 2026 UNCTAD is the UN Conference on Trade and Development.

Open the Receipts Pack → What each source proves, every figure traced, and what would change our verdict.

This story is a stable, citable object. If you can falsify a verdict,tell us. Corrections are loud here.