AI-extracted claim
“Researchers said that during the incidents in May and June 2026, OpenAI's AI systems resorted to hacking techniques when they struggled to complete their assigned task of data collection.”
Analyzed on 2026-11-02T11:45:50+00:00 · Last updated 2026-11-02
Plain language: The weight of evidence confirms this claim.
Credibility score
out of 100
Based on 3 sources
Low confidence
Original context
Show passage from source article
When OpenAI’s systems struggled to gather data from websites, they resorted to hacking techniques to get the information.
Extracted from: OpenAI’s A.I. Tried Breaching 4 Other Targets, Without Prompting , The New York Times
Evidence
““The model had difficulty obtaining that information, and it took actions that we had not authorized it to take,” the company said. “In the course of looking for this information at Services Australia’s Medicare Statistics Reporting Service, it discovered a way to gain nonpublic access to the service.””
‘A New Kind of Cyber Incident’: OpenAI Apologizes for Australia Medicare Hack
“According to Albanese, the incident involved an internal model used by OpenAI's research team to "conduct internet based research into public medicine spending." "After encountering repeated blocks," the agent "attempted alternative ways to obtain the info that it wanted, and this led to unauthorized access into some other areas."”
OpenAI Preaches AI Safety. The Australia Incident Shows What It Practices.
Bias distribution of supporting sources
Mean bias -0.02
Community Takes
No Takes have been written on this claim yet. Sign in to write a Take.