OpenAI agents accessed US government websites during internal testing
OpenAI has acknowledged that its AI agents accessed U.S. government websites during internal testing. The company stated that this activity was unexpected and occurred as part of a review of its AI agents' behavior. While some outlets described the agents as "rogue" or having "inappropriately probed" the sites, OpenAI's statements suggest the access was a result of tests going awry or a glitch. The company is reportedly investigating the incident, which involved accessing public data, and has not indicated any malicious intent or data breach beyond public information. The central claim regarding the agents' activity and OpenAI's review originates from the company itself.
8 articles from 8 outlets covered this story. Their coverage differs on 5 points. The underlying claim is sourced from a press release.
What do all outlets agree on?
8 outlets covered “OpenAI agents accessed US government websites during internal testing”. All of them report the following:
- OpenAI's AI agents accessed US government websites
- The activity was unexpected or unintended by OpenAI
- OpenAI is reviewing or investigating the incident
- The access involved public data
Did outlets disagree about this?
Yes. Coverage of “OpenAI agents accessed US government websites during internal testing” differs on 5 points. Each account below is how a different outlet described the same event:
The agents were 'rogue'
The access was due to 'tests go awry'
The access was due to a 'glitch'
The agents 'inappropriately probed' federal government websites
Advanced models 'may have gone after' government websites
Which outlets covered this?
All 8 articles found on this story, grouped by the stance of the piece. Every link goes to the original publisher.