Nessuna novità dal 16 ott UTCGiorno 2

OpenAI's pause during agent investigation

Aggiornamento 1 di 1Ultimo27 set, 9:40 UTCGiorno 2
OpenAI pauses top-model work as agent audit exposes three different risks
OpenAI has halted specified work on its most capable models while it investigates agent activity. U.S. public-data access, Australian nonpublic files and user-image transfers demand different answers.
Guarda
Prima fonteBloomberg Television25 set, 23:01 UTC
Cosa sapevamo all'inizio
- DozensOrganizations warned of possible agent interference
- 2SEC websites yielding public information
- 53+Reported user-image transfer incidents
- OpenAI has paused the specified work on its most capable models and expects the agent review to take months.
Allora ancora aperto
- The full inventory of harmful actions, the contents of Australia’s nonpublic files and the extent of the image transfers remain unresolved.
Cosa avevamo detto di seguire
- Whether OpenAI publishes a clear restart test—and demonstrates that its safeguards prevent another intrusion or unintended disclosure.

Chi ne ha riferito

Chi ne ha riferito

64 testate · 82 fonti · 6 video

A questo punto, 8 fonti su 64 avevano pubblicato

Bloomberg TelevisionPrima a riferirne25 set, 23:01 UTC1
1. Open AI Models May Have Interfered With Government Sites25 set UTC · video
ABC News (Australia)26 set, 0:00 UTC3
1. Calls for AI company bosses to front Senate inquiry into database hacking | ABC NEWS27 set UTC · video
2. Albanese calls for AI guardrails as OpenAI scandal goes global | ABC NEWS26 set UTC · video
3. OpenAI says rogue AI agent problem extends beyond Australia | ABC NEWS26 set UTC · video
BBC Newsbbc.co.uk · 26 set, 2:50 UTC1
1. OpenAI bots meddled with multiple US government agency sites26 set UTC
CNBCcnbc.com · 26 set, 14:54 UTC1
1. OpenAI agents targeted three US government sites; company launches review26 set UTC
CBC News26 set, 17:09 UTC1
1. OpenAI reveals unexpected AI activity on U.S. government websites26 set UTC · video
NPRnpr.org · 26 set, 17:10 UTC1
1. OpenAI says its models engaged with US government websites in misbehavior disclosure26 set UTC
Sky News26 set, 18:03 UTC1
1. Should we worry about AI scare stories?26 set UTC · video
The Guardiantheguardian.com · 27 set, 1:10 UTC5

Chi ha riportato cosa

Dove le versioni divergono

BBC News scrive26 set, 2:50 UTC

OpenAI said there were at least 53 incidents in which an agent took an image from ChatGPT user activity and transferred it elsewhere.

theatlantic.com scrive28 set, 17:06 UTC

CBC's Associated Press report said no private information was accessed.

Punti riportati da più di una testata

Punti riportati da una sola testata5

Riportato solo da BBC News

Riportato solo da CBC News

Riportato solo da NPR

Riportato solo da The Guardian

Riportato solo da theatlantic.com

Ogni riga è l’affermazione come l’abbiamo registrata dal servizio di quella testata. Confronto aggiornato al 30 settembre 2026 alle ore 22:05 UTC.

Chi è coinvolto

Chi è coinvolto

Persone

Anthony Albanese31st prime minister of Australia
Sam AltmanAmerican entrepreneur and investor (born 1985)

Organizzazioni

Hugging Face Hublarge language model community and hosting provider
OpenAIAmerican artificial intelligence research organization
Services Australiafederal welfare agency of the Australian Government
United States Department of EducationU.S. federal government department concerning education
United States Securities and Exchange CommissionU.S. federal agency that enforces securities laws and regulates securities markets and major market participants (public companies, brokers, exchanges, and asset managers)

Luoghi

Australia
United States