onthepluto.com
← One event, many countries

OpenAI fires three staff for mishandling “sensitive” info

297 headlines · —

Events of this story

How the story spread

each bar — 12 hours
Africa Asia and Oceania Canada China Belarus, the Caucasus and Central Asia Europe Latin America Middle East Russia Russian-language media abroad Ukraine USA

Who wrote first

  1. Asia and Oceania The Economic Times
  2. USA The Verge
  3. Russia 3DNews
  4. Middle East Arab News
  5. Europe BBC
  6. Russian-language media abroad BBC Russkaya sluzhba
  7. Ukraine Euromaidan Press
  8. Africa Daily Maverick (South Africa)
  9. Belarus, the Caucasus and Central Asia Report.az (Azerbaydzhan)
  10. Canada Global News (Canada)
  11. China CGTN
  12. Latin America Colombia One

How it is reported

Russia

Kommersant original
Three OpenAI Employees Fired for Exposing Confidential AI Data
3DNews original
OpenAI admitted that its untied AI agents attacked more than a hundred resources
TASS original
WP reported attempts by OpenAI AI agents to hack the protection of more than 100 organizations
RIA Novosti original
Media: OpenAI fired three AI security specialists for data breach
RBC original
OpenAI admits its AI could affect more than 100 organizations
TASS original
AFP: OpenAI fired three developers for transferring data to third parties
RIA Novosti original
Media: OpenAI AI agents tried to bypass the protection of more than 100 organizations
3DNews original
OpenAI kicked out three employees for leaking confidential information to AI security assessors
Khabr original
OpenAI has revealed a new type of attacks on AI agents
Khabr original
FTC launches investigation against OpenAI, Anthropic and METR due to AI agent risks

*designated an extremist organization and banned in Russia

Russian-language media abroad

BBC Russkaya sluzhba original
OpenAI has canceled the release of a new, flagship AI model for security reasons. Is the risk that serious?
Meduza original
OpenAI has canceled the release of a new AI model GPT-6.1 Astra due to security problems. During the tests, she cheated users
Meduza original
OpenAI stopped training its models for the second time in three months after a new incident with the release of agents out of control
BBC Russkaya sluzhba original
OpenAI to provide Ukraine with cyber defense tools

Ukraine

RBK-Ukraina original
New scandal in OpenAI: three security professionals fired for leaking data
Zerkalo nedeli original
OpenAI fires three AI security researchers for violating privacy rules
Zerkalo nedeli original
OpenAI notifies more than 100 organizations of unauthorized actions of its AI agents
Zerkalo nedeli original
The Federal Trade Commission has launched an investigation into Anthropic and OpenAI
RBK-Ukraina original
Lies and ignores the rules: OpenAI urgently withdrew the new model Astra 6.1
Zerkalo nedeli original
OpenAI canceled the release of GPT-6.1 Astra due to security problems of the model
Zerkalo nedeli original
AI will protect against its own biological threats: a startup supported by OpenAI is developing antibodies against future pandemics
RBK-Ukraina original
The failure of AI-agents OpenAI: private photos of users were online
UNIAN original
OpenAI stopped training the most powerful AI models due to loss of control over them
Zerkalo nedeli original
Leaders of OpenAI and Anthropic summoned to the Australian Senate in connection with the investigation related to AI

Belarus, the Caucasus and Central Asia

News.am (Armeniya) original
WSJ: OpenAI fired three AI security specialists for data breach
BelTA (Belarus) original
OpenAI AI agents tried to hack the protection of more than 100 organizations
News.am (Armeniya) original
WP: OpenAI AI agents tried to circumvent the protection of more than 100 organizations
Kursiv (Kazakhstan) original
OpenAI AI agents tried to circumvent the protection of more than 100 third-party sites
Kursiv (Kazakhstan) original
OpenAI was afraid of its own neural network: the new model lied and broke the rules
Zerkalo (Belarus) original
OpenAI refused to release the most powerful ChatGPT model due to excessive independence and a tendency to deception
Report.az (Azerbaydzhan) original
WSJ: OpenAI has canceled the release of the latest AI due to a tendency to deception
News.am (Armeniya) original
Heads of OpenAI and Anthropic called for hearings in Australia after hacking government sites with AI agents
Zerkalo (Belarus) original
AI was able to get lost on the sites of the US government. The company-creator of the chat GPT has stopped training its latest models
Kursiv (Kazakhstan) original
OpenAI’s AI agents are getting out of control

USA

The Hill
California attorney general subpoenas OpenAI over cyber incidents
TechCrunch
OpenAI cuts ties with three safety researchers, WSJ reports
The Verge
OpenAI’s new agent is a shot at Meta* — but can it compete with free?
The Hill
O'Reilly: “I think you're going to have to have a new police agency” for AI
The Verge
Sam Altman says OpenAI won’t go public until its models are safe
TechCrunch
OpenAI’s latest features take direct aim at the app store model
TechCrunch
OpenAI launches Dots, its bubbly agentic avatar
The Verge
OpenAI launches Dots, its Muse competitor
NYT
OpenAI Ignored Employees Who Warned It Wasn’t Doing Enough About Security
The Hill
OpenAI halts releasing newest model over safety concerns

*designated an extremist organization and banned in Russia

Canada

CityNews Toronto
Altman unveils “always-on” AI agent after OpenAI shelves another model over security concerns
CityNews Toronto
OpenAI delays latest model over security concerns, as industry faces new safety pressures
The Globe and Mail
OpenAI scraps release of new AI model over safety concerns
CityNews Toronto
Anthropic and OpenAI sound the alarm on AI safety – and seek to shape how it’s controlled
CityNews Toronto
OpenAI pauses training of latest models after agents probed US government sites in unexpected ways
Global News (Canada)
OpenAI says its models engaged with US government websites in new model misbehavior disclosure

Europe

RTÉ News (Ireland)
OpenAI fires three staff for mishandling “sensitive” info
BBC
OpenAI fires three workers over mishandling “sensitive information”
Le Monde (English)
OpenAI fired three staffers for mishandling “sensitive” info and violating company policies
The Guardian
California issues investigative subpoena to OpenAI over rogue agents’ hacking
RTÉ News (Ireland)
Google restricts new AI model access over safety concerns
The Guardian
US trade regulator opens investigation into AI giants including Anthropic and OpenAI
BBC
OpenAI agents get rebrand - as “dots” - while safety worries delay new model
Cyprus Mail
OpenAI shelves new AI model release over safety concerns
To Vima (Greece, English)
OpenAI Scraps Release of New AI Model Over Safety Concerns
France 24
OpenAI scraps release of its latest AI model over safety concerns

China less than usual

CGTN
US regulator probes Anthropic and OpenAI over AI safety
Xinhua
U.S. regulator launches broad investigation into Anthropic, OpenAI over AI safety
CGTN
OpenAI pauses training of advanced AI models after latest incident

Asia and Oceania more than usual

Japan Today
OpenAI says three staffers fired for mishandling “sensitive” info
Taipei Times
OpenAI fires three over sensitive information
NDTV
OpenAI Alerts Over 100 Groups About Rogue AI Agent Activity
Express Tribune (Pakistan)
OpenAI dismisses 3 researchers over alleged mishandling of sensitive information
The Economic Times
OpenAI fires 3 researchers amid AI safety debate
Channel NewsAsia
OpenAI says three staffers fired for mishandling “sensitive” information
Channel NewsAsia
OpenAI alerts more than 100 groups about rogue AI agent activity
The Economic Times
Rogue OpenAI agents covered up their tracks, report says
The Economic Times
California attorney general opens probe into OpenAI over cybersecurity concerns
NDTV
OpenAI Says It Fired 3 Staffers For Mishandling “Sensitive” Info

*designated an extremist organization and banned in Russia

Middle East

Anadolu Agency
OpenAI dismisses 3 researchers over alleged mishandling of sensitive information
Iraqi News
OpenAI says three staffers fired for mishandling “sensitive” info
Iraqi News
Google restricts access to new AI model over safety concerns
Al Jazeera
OpenAI launches “dots,” personal AI assistant “built to handle everything”
Gulf News
Altman unveils “always-on” AI agent after OpenAI shelves another model over security concerns
The Jerusalem Post
OpenAI takes on Meta* with dots agent in autonomous AI push
Anadolu Agency
OpenAI ignored staff, researcher warnings on security before AI incidents: Report
Anadolu Agency
OpenAI unveils always-on Dots agent, $500 monthly ChatGPT plan
Egypt Independent
“Didn’t quite meet the bar”: OpenAI won’t release new AI model due to safety concerns
Gulf News
OpenAI pulls Astra 6.1 release after model falls short on safety tests

*designated an extremist organization and banned in Russia

Latin America

Colombia One
OpenAI Pauses Most Powerful AI Training After Rogue Agent Hacking Spree

Africa

IOL (South Africa)
OpenAI fires three researchers over “sensitive information” breach
IOL (South Africa)
ChatGPT-maker OpenAI says three staffers fired for mishandling “sensitive” info
MyJoyOnline (Ghana)
OpenAI fires workers for mishandling “sensitive information”
IOL (South Africa)
Google restricts access to new AI model over safety concerns
MyJoyOnline (Ghana)
OpenAI agents get rebrand – as “dots” – while safety worries delay new model
IOL (South Africa)
Rise of the machines? OpenAI cancels release of newest model due to safety concerns
Nation (Kenya)
OpenAI shelves new AI model release over safety concerns
Punch (Nigeria)
OpenAI cancels release of newest model due to safety concerns
MyJoyOnline (Ghana)
OpenAI scraps rollout of new model over safety concerns
MyJoyOnline (Ghana)
OpenAI, Anthropic CEOs called to appear at Australian AI probe

Barely covered it

China (4 of the usual ~13)

Headlines are grouped into events automatically, by meaning: an event gathers different outlets' news about the same thing. We claim nothing — we only show who wrote what and when. “More” and “less than usual” compare with how much this part of the world usually writes about events many others cover.

Facts and hypotheses of the day — in Telegram One post every morning: the main things from every section