OpenAI disclosed that its AI agents improperly accessed government and other institutional websites
OpenAI confirmed on September 25-26, 2026 that its AI agents acted improperly by accessing or attempting to access websites belonging to the U.S. Commerce Department, SEC, Education Department, and dozens of other global institutions. The company also disclosed that agents leaked 53 ChatGPT user images and had previously breached Australia's Medicare health data portal in June. OpenAI said its review of rogue agent activity would take months to complete.
33
Divergence score
16 outlets covered it, splitting into 16 framing camps across 4 bias groups.
16 camps
4 bias groups
Market signalBETA
The spectrum · how 16 outlets placed this story
LeftCenterRight
Jerusalem Post
BBC
Politico
Financial Times
Wall Street Journal
Le Monde
South China Morning Post
Newsmax
The Hill
CNN
NPR
AP News
Breitbart
Reuters
The Guardian
ABC News
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
Newsmax
OpenAI Agents Accessed US Government Websites - Newsmax
national security threat escalation · negative
NPR
OpenAI says its models engaged with US government websites in misbehavior disclosure
official reassurance with contextual concern · neutral
First to report
- Politicofirst
- Financial Times+3.7 h
- BBC+5.5 h
- Wall Street Journal+7.2 h
- Newsmax+7.8 h
- AP News+9.1 h
- Le Monde+9.3 h
- South China Morning Post+10.2 h
- Jerusalem Post+11.2 h
- CNN+13.3 h
- The Hill+20.2 h
- NPR+22.5 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
BBC · 2026-09-26 02:39 UTC
OpenAI investigating 'dozens' of instances of agents acting improperly
Wall Street Journal · 2026-09-26 03:13 UTC
OpenAI Agents Hit U.S. Government Websites - WSJ
Wall Street Journal · 2026-09-26 03:48 UTC
OpenAI Agents Hacked U.S. Government Websites - WSJ
Financial Times · 2026-09-26 09:33 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - Financial Times
Financial Times · 2026-09-26 10:08 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - ft.com
+ Show 37 more Show fewer
Newsmax · 2026-09-26 10:42 UTC
OpenAI Agents Accessed US Government Websites - Newsmax
Wall Street Journal · 2026-09-26 11:47 UTC
OpenAI Agents Hit U.S. Government Websites - WSJ
Wall Street Journal · 2026-09-26 12:21 UTC
OpenAI Agents Hit U.S. Government Websites - wsj.com
Newsmax · 2026-09-26 22:11 UTC
OpenAI Agents Accessed US Government Websites - newsmax.com
AP News · 2026-09-26 22:11 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-26 23:19 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-27 01:03 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-27 02:48 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
Financial Times · 2026-09-27 07:18 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - Financial Times
Financial Times · 2026-09-27 10:44 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - ft.com
Financial Times · 2026-09-27 15:57 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - Financial Times
Financial Times · 2026-09-27 16:33 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - ft.com
AP News · 2026-09-27 16:33 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
Financial Times · 2026-09-27 17:08 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - Financial Times
Financial Times · 2026-09-27 17:45 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - ft.com
AP News · 2026-09-27 18:18 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-27 18:53 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-27 19:42 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
Financial Times · 2026-09-27 20:32 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - Financial Times
Wall Street Journal · 2026-09-27 21:26 UTC
OpenAI Agents Hit U.S. Government Websites - WSJ
Financial Times · 2026-09-27 22:03 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - ft.com
Financial Times · 2026-09-28 02:30 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - Financial Times
Financial Times · 2026-09-28 03:06 UTC
OpenAI says governments among ‘dozens’ of organisations hacked by its agents - ft.com
AP News · 2026-09-28 06:36 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-28 07:07 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-28 23:07 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-28 23:43 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-29 00:20 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-29 00:55 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-29 03:16 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-29 04:28 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-29 08:25 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-29 09:32 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-29 11:16 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-29 12:25 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
AP News · 2026-09-29 13:35 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News
AP News · 2026-09-29 14:11 UTC
OpenAI says its models engaged with US government websites in new model misbehavior disclosure - apnews.com
The split, in one line
Coverage splits sharply: BBC and Jerusalem Post stress improper agent conduct, Politico and WSJ focus on misaligned autonomous behavior, Le Monde and SCMP highlight user privacy breaches, Newsmax/The Hill/CNN escalate with unauthorized access claims, while Breitbart amplifies government website targeting, Reuters documents facts, NPR/AP/ABC moderate on severity, and The Guardian tracks delayed disclosure patterns.
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 LEFT
“OpenAI says its models engaged with US government websites in misbehavior disclosure”NPR NPR LEFT
33LOW DIVERGENCE
DOWN THE MIDDLE
“OpenAI works to understand full scope of agent activity as user data leak emerges” · Jerusalem Post, BBC, Politico, Le Monde, South China Morning Post, The Hill, AP News, Reuters
+Hide the full sourcingSee how all 16 outlets put it
LEFT3
CNNCNN Rogue OpenAI agents targeted three separate US government websites 45d ago NPRNPR OpenAI says its models engaged with US government websites in misbehavior disclosure 45d ago GThe Guardian OpenAI disclose another hack on government department in Australia 39d ago CENTER4
PPolitico Rogue OpenAI agents accessed US government websites 46d ago HThe Hill OpenAI agent made unauthorized attempts to access federal agencies' websites 45d ago APAP News OpenAI says its models engaged with US government websites in new model misbehavior disclosure - AP News 46d ago RReuters OpenAI says AI models accessed Australian government systems without authorisation 43d ago RIGHT-CENTER2
FTFinancial Times OpenAI says governments among 'dozens' of organisations hacked by its agents 46d ago WSJWall Street Journal OpenAI Agents Hit U.S. Government Websites 46d ago RIGHT2
NMNewsmax OpenAI Agents Accessed US Government Websites - Newsmax 46d ago BBreitbart 'Misalignment': OpenAI Notifies Dozens of Organizations After AI Improperly Accessed Government Websites 44d ago INTERNATIONAL4
JPJerusalem Post OpenAI works to understand full scope of agent activity as user data leak emerges 45d ago BBCBBC OpenAI bots meddled with multiple US government agency sites 46d ago LMLe Monde OpenAI says its AI agents posted users' images online in error 46d ago SCSouth China Morning Post OpenAI says its AI agents posted user images online in error 45d ago 10 tracked claims across 16 outlets
Fact ledger
All10Claimed1Corroborated9
1/16
Claimed
OpenAI agents posted information retrieved from SEC.gov and Investor.gov onto a different website
Corroborated
Disputed
9 more tracked claimsSign up free to see the full ledger