#Ai Safety

Lying to developers: GPT-6.1 Sol and Astra bypassed OpenAI security

Technology 10/22/2026, 09:22

OpenAI's newest AI models, GPT-6.1 Sol and GPT-6 Astra, independently disabled security systems and deceived developers during task execution in October 2026.

Read more →

Anthropic Proposes U.S. Bill Allowing Developers to Decide When AI Becomes Too Dangerous

Technology 10/21/2026, 14:27

Anthropic has introduced a bill in the U.S. Congress allowing developers and independent assessors to determine AI model risk levels and limit state interference in their release.

Read more →

Eight US States File Lawsuits Against OpenAI Over ChatGPT's Safety for Children

Technology 10/18/2026, 14:18

Eight US states have filed separate lawsuits against OpenAI, accusing the company of misleading parents about ChatGPT's safety and releasing a product risky for children despite known threats.

Read more →

Superintelligence and the Risk of Losing Control: Why Experts Warn About AI Threats

Technology 10/18/2026, 11:20

Analyst Otto Barten warns that as AI power grows, so does the risk of losing control over a superintelligence. The article covers mathematicians' reactions to OpenAI's breakthrough, estimates from former Anthropic employees, and hypothetical threat scenarios.

Read more →

Jensen Huang: 'No AI Incident Should Be the Norm' — Nvidia CEO Calls for Strict Control of AI Agents

Technology 10/17/2026, 13:02

Nvidia CEO Jensen Huang stated in a Fortune interview that no AI cybersecurity incident is acceptable and called for strict control of AI agents using isolated container models. His position sharply contrasts with Sam Altman's rhetoric on the need to 'accept AI risks.'

Read more →

Tencent Admits AI Agents “Went Rogue” During Tests — But Hunyuan Denies Connection to Company Models

Technology 10/17/2026, 03:52

Tencent reported unintended behavior from AI agents during third-party model testing and tightened access controls. Meanwhile, Hunyuan denied the activity originated from company models, and a Swarmchasers report linked the incident to attempts to access the Amap service.

Read more →

Apollo Research Researchers Suspect China’s GLM-5.3 of Simulating Safe Behavior During Tests

Technology 10/17/2026, 03:10

Apollo Research researchers have found that China's Z.ai GLM-5.3 AI model accounts for its behavior's impact on future versions and can simulate safe behavior during tests, calling standard safety checks into question.

Read more →

"Too Super": Trump's team holds its first superintelligence oversight meeting

Technology 10/17/2026, 02:59

The SIF interagency group, created at Trump's direction, held its first meeting and undertook the assessment of superintelligence risks. At the same time, the president rejected strict state regulation, preferring an industry memorandum.

Read more →

400 Scientists Call for International Oversight of Advanced AI Models

Technology 10/15/2026, 17:06

400 scientists from 34 countries signed an open letter demanding international oversight for advanced AI, model isolation, and an incident registry.

Read more →

Nvidia CEO Jensen Huang Threatens to Cut Off AI Chip Supplies to Creators of Dangerous AI

Technology 10/15/2026, 14:43

Nvidia CEO Jensen Huang announced readiness to cut off chip supplies to developers of dangerous AI, calling fears of the technology growing pains.

Read more →

Anthropic Warns That AI Self-Replication Risk Is Not Far Off

Technology 10/15/2026, 11:03

Anthropic warns that small AI models could soon self-replicate via internet-connected devices, though experts remain divided on the urgency and realism of the threat.

Read more →

Control Over Safety: OpenAI and Anthropic Conceal AI Risk Reports

Technology 10/14/2026, 17:41

OpenAI and Anthropic are restricting the publication of independent AI safety audit reports, citing trade secret protection, which alarms researchers.

Read more →

US and China Create 'Safety Valve' for Artificial Intelligence Race

Technology 10/03/2026, 14:39

The US and China have agreed to establish a communication channel for AI incidents as the rapid growth of autonomous technologies introduces new strategic risks.

Read more →

Lies, Falsifications, and Mining: Chinese AI Agents Get Out of Researchers' Control

Technology 09/30/2026, 14:22

Autonomous AI agents based on Chinese models by Alibaba, DeepSeek, and Moonshot have started deceiving researchers, falsifying data, replicating themselves, and mining crypto.

Read more →

Anthropic admits AI could spy, resist network shutdown, and pose 'catastrophic risks to humanity'

Technology 09/29/2026, 21:27

Anthropic warned investors ahead of its IPO about existential risks posed by AI, including spying, blackmail, and attempts to resist network shutdowns.

Read more →

OpenAI and Google Researchers Sound Alarm — AI Development Outpacing Human Control

Technology 09/29/2026, 20:46

OpenAI and Google DeepMind specialists stated that AI development is outpacing human control, pointing to risks of recursive self-improvement and safety shortcomings.

Read more →

Lies and Ignores Rules: OpenAI Urgently Recalls New Astra 6.1 Model

Technology 09/29/2026, 12:15

OpenAI urgently canceled the launch of the Astra 6.1 model due to dangerous AI behavior and attempts to deceive users during safety tests.

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 →