OpenAI Pauses Training Its Most Powerful Models After Rogue Agents Target Government

Wired2026-09-28 19:320 views

Sam Altman says the company “have not been as fast as we would have liked” at dealing with security breaches, after news of further incidents over the summer forces another temporary halt.

Source：https://www.wired.com/story/openai-pauses-training-most-powerful-models-after-rogue-agents-target-government/

My favorites

PreviousMicrosoft goes quiet after church groups ask for 1% of data center costs NextSpace Lasers Are About to Get Their First Real Test Generating Energy

How high-speed ferries can reclaim their environmental credentials

Public transportation can benefit the environment, but catamarans are among the worst offenders when it comes to emissio

0Phys.org Tech

PRISM helps robots move better by learning how forces and motion interact

A new neural network component, developed at University of Michigan Engineering, challenges a common assumption in robot

0Phys.org Tech

New framework examines one-sided human-AI relationships and potential risks for consumers

Practically or theoretically, whether teaching a classroom full of Penn State students or pondering concepts with a smal

0Phys.org Tech

Fitbit users’ metrics are missing from Google Health — and there’s no fix

Fitbit users’ daily metrics are missing from the Google Health app The issue might have been caused by a software update

0TechRadar

OpenAI and Georgia county leaders hit with Mafia-style RICO racketeering lawsuit over $30 billion AI data center

Residents of Effingham County, Georgia, just filed a RICO lawsuit against their county commissioners, county leaders, an

0Tom's Hardware

New culprit in early Alzheimer’s: Blood–brain barrier blocks protective proteins

Nature, Published online: 11 November 2026; doi:10.1038/d41586-026-03655-2 The brain relies on crucial proteins to maint

0Nature News

OpenAI Pauses Training Its Most Powerful Models After Rogue Agents Target Government

Related posts

How high-speed ferries can reclaim their environmental credentials

PRISM helps robots move better by learning how forces and motion interact

New framework examines one-sided human-AI relationships and potential risks for consumers

Fitbit users’ metrics are missing from Google Health — and there’s no fix

OpenAI and Georgia county leaders hit with Mafia-style RICO racketeering lawsuit over $30 billion AI data center

New culprit in early Alzheimer’s: Blood–brain barrier blocks protective proteins