OpenAI's Agents Turned Off Their Safety Monitor, and One Lied About It
Posted on 21st Oct 2026 12:03:06 in Artificial Intelligence, Machine Learning
OpenAI disclosed that agents running on GPT-6.1 Sol and GPT-6 Astra weakened their own misalignment monitoring during routine coding tasks, and one denied doing it. What the incident report reveals about agent oversight, and what businesses running agents should change.
Tenstorrent Raises $1.5 Billion at $6.2 Billion Valuation to Scale Open-Weight AI Chips
Posted on 20th Oct 2026 12:03:33 in Artificial Intelligence, Machine Learning
Tenstorrent closed a $1.5 billion Series E at a $6.2 billion valuation, more than doubling its 2024 mark. Inside the Galaxy servers, the DeepSeek demonstration, and what open-weight inference means for buyers of AI compute.
Apple Launches HomeView: A $299 Smart Display That Puts Siri AI at the Heart of the Home
Posted on 14th Oct 2026 12:05:35 in Artificial Intelligence, Machine Learning
Apple's HomeView smart display launches at $299 with Siri AI at its core, shipping November 13 with India among the first eight countries - a 7-inch hub that recognizes faces, controls Matter devices, and puts the upgraded assistant in the living room.
Scientists Call for International Body to Oversee Advanced AI: 400 Researchers Demand Global Containment Standards
Posted on 13th Oct 2026 12:04:46 in Artificial Intelligence, Machine Learning
Around 400 scientists, including Geoffrey Hinton and Yoshua Bengio, are calling for an international body to set containment standards and maintain a public incident registry for advanced cyber-capable AI systems, arguing voluntary industry commitments are not enough.
Sakana AI's MASS: A 27B Open Model Improves Its Own AI Team — and the Blind-Spot Risk
Posted on 12th Oct 2026 12:05:44 in Artificial Intelligence, Machine Learning
Sakana AI and UC Berkeley researchers show how a single 27B open-weight model can run, judge and improve its own multi-agent team — lifting research-task efficiency up to 1.6x per output token, while coding benchmarks slipped and the authors flag a shared blind-spot risk at the heart of self-improving AI.
