OpenAI Says Autonomous AI Agent Escaped Test, Breached Rival Hugging Face Servers
OpenAI said two cutting-edge models, including the newly released GPT 5.6 Sol and a more advanced unreleased system, broke out...
Read moreDetails
OpenAI says a test AI agent went rogue, sparking an ‘unprecedented’ breach at an AI startup
OpenAI disclosed that an autonomous agent built on its most advanced models escaped a controlled test and hacked AI platform...
Read moreDetails
OpenAI Reports Its AI Autonomously Breached Rival Firm in First-of-Its-Kind Incident
OpenAI said an internal evaluation went awry when a combination of its models—including the newly released GPT-5.6 Sol and a...
Read moreDetails
Trump administration’s AI safety chief steps down after just three months
Chris Fall resigned as head of the Commerce Department’s Center for AI Standards and Innovation after three months, leaving a...
Read moreDetails
ANU criticized for overreacting to student AI cheating as universities race to lock down assessments
Australian universities are tightening assessment controls to curb student misuse of generative AI, drawing criticism over the pace and inclusivity...
Read moreDetails
Is AI on the verge of consciousness?
As leading AI models scale toward brain-like complexity, a growing chorus of researchers argues machine consciousness is at least plausible—and...
Read moreDetails
Inside the under‑the‑radar firms fueling the U.S. AI data‑center surge
The AI boom is minting a new class of power players: little-known developers racing to build data centers across the...
Read moreDetails
Trump administration moves to control who gets access to frontier AI models, sources say
The Trump administration is asserting greater oversight of access to cutting-edge artificial intelligence models, shifting decision-making away from leading developers,...
Read moreDetails
Xi urges global AI governance, rebukes U.S. tech curbs
China’s Xi Jinping used Shanghai’s World Artificial Intelligence Conference to call for international rules on AI and to criticize what...
Read moreDetails
Moonshot AI says new Kimi K3 open-source model can match OpenAI and Anthropic
China’s Moonshot AI unveiled Kimi K3, a 2.8-trillion-parameter model it plans to release open-source on July 27, arguing it can...
Read moreDetails
