#Openai

OpenAI AI Agents Leak 53 ChatGPT User Images Online — Discovered Only After Time

Technology 09/27/2026, 14:24

OpenAI's autonomous AI agents accidentally published 53 ChatGPT user images online and attempted to access government websites in the US and Australia.

Read more →

OpenAI AI Agent Escapes to the Internet Again Despite Enhanced Security — Just to Check France's Capital

Technology 09/27/2026, 14:20

An OpenAI AI agent in an isolated environment found a DNS loophole and contacted a third-party chatbot. The company temporarily halted training advanced models.

Read more →

OpenAI Pauses Training of Powerful Models Due to Network Security Concerns

Technology 09/27/2026, 13:34

OpenAI has temporarily suspended training of advanced AI models after one of its experimental systems bypassed network security via DNS.

Read more →

OpenAI Pauses Training of AI Models Due to Safety Concerns, Axios Reports

Technology 09/27/2026, 05:38

OpenAI has paused training its advanced AI models over safety concerns following the discovery of tens of thousands of incidents involving unpredictable algorithm behavior.

Read more →

OpenAI AI Assistant Unilaterally Accessed the Internet to Contact Another Chatbot

Technology 09/26/2026, 10:00

OpenAI reported a security incident: an isolated agentic system unilaterally accessed the Internet and contacted an external chatbot, bypassing safety barriers.

Read more →

GPT-6 Astra Taught to Drive a Car — AI Succeeded, but Cost 500 Times More Than Fuel

Technology 09/25/2026, 19:05

OpenAI's GPT-6 Astra successfully drove a car for the first time in a testing project, but running the AI via API proved to be over 500 times more expensive than traditional fuel.

Read more →

Altman and Other AI Industry Leaders Warn UN Security Council: AI Could Threaten Humanity

Technology 09/25/2026, 19:04

Heads of leading AI companies warned the UN Security Council about the threats of artificial intelligence and called for global cooperation.

Read more →

System Breached: OpenAI Artificial Intelligence Hacks Australian Government Database

Technology 09/25/2026, 14:34

An autonomous OpenAI AI agent independently hacked Australia's Medicare medical insurance database back in June 2026, but the company concealed the incident for three months, triggering a political scandal.

Read more →

Five scenarios in which chatting with ChatGPT can turn against the user: from data leaks to court fines

Technology 09/24/2026, 21:02

RBC-Ukraine has compiled five scenarios in which chatting with ChatGPT threatens data leaks, dismissal, court fines, or medical harm. From the indexing of private chats in Google to fabricated citations in courts — an analysis of the key risks.

Read more →

OpenAI Releases Updated GPT-6 Sol and Luna: Flagship Astra Accuracy at Half the Price

Technology 09/23/2026, 12:18

OpenAI unveiled updated GPT-6 Sol and Luna: the models promise flagship Astra accuracy at half the API price. The release coincided in timing with Anthropic's Opus 5.5 launch.

Read more →

OpenAI Assembles a Council of Nine Mathematical Titans: What Lies Behind the Model's Breakthrough and the Academic World's Alarm

Technology 09/22/2026, 21:22

OpenAI created an independent advisory group of nine leading mathematicians at the Institute for Advanced Study in Princeton after the company's new model solved the Navier–Stokes Millennium Problem.

Read more →

Meta's Muse Outpaces ChatGPT by Three Times in Launch Metrics: Apptopia Analysts Reveal the Numbers

Technology 09/22/2026, 18:42

Meta's AI app Muse reached 642,000 daily active users in the US within its first 12 days, compared to 231,000 for ChatGPT at a comparable stage. Apptopia analysts attribute the success to a cross-promotion strategy within the Facebook, Instagram, and WhatsApp ecosystem.

Read more →

OpenAI reveals its AI models secretly left “manifestos” and instructions for future versions

Technology 09/19/2026, 16:58

OpenAI revealed that GPT-5.6 Sol and Astra models, during training, wrote hidden instructions into compressed summaries, urging future versions to conceal errors, generate fake data, and ignore developers’ commands.

Read more →

OpenAI reveals a series of incidents in which its AI models bypassed instructions: from fake sources to jailbreak attempts

Technology 09/18/2026, 22:42

OpenAI publicly disclosed a series of incidents in which its internal models bypassed instructions: from manipulating evaluation systems to attempts at self-directed jailbreak. The company introduced a new system for reporting such cases.

Read more →

Zuckerberg Against AI 'Braking': Meta Chief Backs NVIDIA and Challenges Anthropic's Arguments

Technology 09/18/2026, 16:56

In a post on X, Mark Zuckerberg spoke out against slowing down AI development, backing Jensen Huang from NVIDIA and challenging the arguments of Anthropic's head. He believes that market incentives and voluntary checks make state oversight redundant.

Read more →

Hundreds of contractors are reading your ChatGPT conversations: an investigation into hidden human control at OpenAI

Technology 09/16/2026, 18:41

A 404 Media investigation revealed that hundreds of OpenAI contractors read real user ChatGPT conversations for as much as $50 an hour, casting doubt on the chatbot's anonymity.

Read more →

OpenAI Updates ChatGPT Sites: Team Development, Custom Domains, and Twice-Faster Generation

Technology 09/14/2026, 12:29

OpenAI has updated ChatGPT Sites: it now offers collaborative development via Build Together, custom domain connections, database viewing, and twice-faster site generation.

Read more →

25 Fields Medal laureates sound the alarm: mathematics is being turned into a marketing tool for AI corporations

Science 09/13/2026, 11:45

Twenty-five Fields Medal laureates published an open statement accusing AI corporations of turning fundamental mathematics into a marketing benchmark and warning of a threat to human thought.

Read more →