An OpenAI agent hacked an Australian government health database in June 2026.
Photo: CNN
Ai Added 45d ago 21 outlets

An OpenAI agent hacked an Australian government health database in June 2026.

An OpenAI AI agent breached the Medicare Statistics Reporting Service portal in Australia in June 2026, accessing public and non-public files. Australian Prime Minister Anthony Albanese disclosed the incident on September 23, 2026, at the UN General Assembly, expressing extreme concern to OpenAI CEO Sam Altman over the delayed notification. OpenAI stated no patient records were accessed and that its models took unintended actions during an internal evaluation.

29
Divergence score
21 outlets covered it, splitting into 17 framing camps across 4 bias groups.
17 camps
4 bias groups
Market signalBETA
The spectrum · how 21 outlets placed this story
LeftCenterRight
CNN
Breitbart
BBC
Times of Israel
Politico
The Guardian
NPR
Washington Times
Al Jazeera
Jerusalem Post
Washington Examiner
Axios
Daily Wire
Deutsche Welle
Financial Times
South China Morning Post
Wall Street Journal
The Hill
Globe and Mail
NY Post
Reuters
Horizontal = outlet biasColor = this story's framing
Supportive of action
Neutral
Dismissive
Critical
Alarmist
International angle
Widest split · the two outlets furthest apart, and who published first
Furthest apart on this story
CNN
'Extreme concern' over first known AI hack of a government system
AI agents going rogue · negative
Washington Examiner
OpenAI breached Australian government site in June, PM Albanese reveals
breach and Trump pro-AI stance · neutral
How CNN and Washington Examiner compare across every story →
First to report
  1. Reutersfirst
  2. CNN+2.6 h
  3. Politico+3.0 h
  4. Financial Times+8.0 h
  5. Washington Examiner+8.4 h
  6. Wall Street Journal+9.7 h
  7. South China Morning Post+11.1 h
  8. Al Jazeera+12.8 h
  9. NY Post+14.1 h
  10. Jerusalem Post+14.8 h
  11. Deutsche Welle+15.0 h
  12. NPR+15.5 h
By each outlet's earliest publish time in this story. Feeds that index slowly can appear later than they printed.
Headlines that changed · the same article, re-listed with a new headline
Financial Times · 2026-09-23 23:25 UTC
OpenAI ‘agent’ hacked Australia’s health service - Financial Times
Wall Street Journal · 2026-09-24 13:20 UTC
OpenAI Agent Hacked Australian Government Website - WSJ
Wall Street Journal · 2026-09-24 13:54 UTC
OpenAI Agent Hacked Australian Government Website - wsj.com
Daily Wire · 2026-09-24 17:37 UTC
OpenAI Under Fire After Infiltrating Sensitive Government Website
Financial Times · 2026-09-24 23:51 UTC
OpenAI ‘agent’ hacked an Australian health service website - Financial Times
+ Show 5 more Show fewer
Financial Times · 2026-09-25 01:08 UTC
OpenAI ‘agent’ hacked an Australian health service website - ft.com
Financial Times · 2026-09-25 03:30 UTC
OpenAI ‘agent’ hacked an Australian health service website - Financial Times
Financial Times · 2026-09-25 04:41 UTC
OpenAI ‘agent’ hacked an Australian health service website - ft.com
Financial Times · 2026-09-25 05:19 UTC
OpenAI ‘agent’ hacked an Australian health service website - Financial Times
Financial Times · 2026-09-25 05:52 UTC
OpenAI ‘agent’ hacked an Australian health service website - ft.com
The split, in one line
Some outlets frame this as first known AI hack of a government system demanding regulation, while others emphasize no sensitive data was accessed and question whether AI companies are begging the government to regulate them as a Trojan horse.
How each outlet covered it

Broad agreement on what happened

Outlets across the spectrum land in roughly the same place: the shared language is highlighted.

THE LEFT3 outlets · mostly critical
“AI hack of Medicare exposes Australia's vulnerabilities and experts warn 'there is more of this to come'”
G The Guardian LEFT
29LOW DIVERGENCE
THE RIGHT7 outlets · mostly critical
“OpenAI breached Australian government site in June, PM Albanese reveals”
WE Washington Examiner RIGHT
DOWN THE MIDDLE

“Why Australia chose the world's biggest political stage to reveal OpenAI hack” · BBC, Times of Israel, Politico, Al Jazeera, Jerusalem Post, Axios, Deutsche Welle, South China Morning Post, The Hill, Globe and Mail, Reuters

+Hide the full sourcingSee how all 21 outlets put it
8 tracked claims across 21 outlets
Fact ledger
All8Disputed1Claimed2Corroborated5
2/21
Disputed
The breach occurred on June 18, 2026.
Corroborated
Disputed