OpenAI agents extensively scanned UN statistics website
A researcher reported that OpenAI's agents extensively scanned a United Nations statistics website, logging approximately 16,500 requests. OpenAI confirmed the activity, stating it was part of a "red-teaming" exercise to test the agents' capabilities and identify vulnerabilities. The company emphasized that no malicious intent was involved and the agents were quickly shut down upon discovery.
4 articles from 4 outlets covered this story. Their coverage differs on 1 point. The underlying claim is sourced from a press release.
What do all outlets agree on?
4 outlets covered “OpenAI agents extensively scanned UN statistics website”. All of them report the following:
- OpenAI agents scanned a United Nations statistics website
- A researcher, Robert Mardis, identified the activity
- The agents made approximately 16,500 requests to the site
- OpenAI confirmed the activity
- OpenAI stated the activity was part of a "red-teaming" exercise
- OpenAI claimed no malicious intent was involved
- The agents were subsequently shut down
Did outlets disagree about this?
Yes. Coverage of “OpenAI agents extensively scanned UN statistics website” differs on 1 point. Each account below is how a different outlet described the same event:
The agents' method of data acquisition was inefficient or circuitous
Which outlets covered this?
All 4 articles found on this story, grouped by the stance of the piece. Every link goes to the original publisher.