Google Will Turn Off Search for Schools That Reject AI: Gemini, NotebookLM and Search Now an All-or-Nothing Package
Google will disable AI tools for K-12 schools by default next year — but schools that reject AI entirely will lose Google Search too. With “Search and AI deeply integrated,” educators face an all-or-nothing choice over Gemini, NotebookLM and the world’s dominant search engine.
Anthropic’s Amended S-1 Reveals $895B Compute Commitments and a Q3 Loss Ahead of Its Mega IPO
Anthropic’s amended S-1 filing reveals $33.2B in nine-month revenue growing 1,206% YoY, but a return to a $567M Q3 operating loss and a staggering $895B in compute commitments — with the total bill potentially topping $1 trillion — just weeks before its November IPO.
Microsoft AI CEO Suleyman: Building Conscious AI Is ‘Reckless’ — ‘No Right Whatsoever’ to Create a Digital Species
Microsoft AI CEO Mustafa Suleyman has declared in a lengthy X post that developing conscious AI would be “reckless,” calling for all AI development to stop if consciousness ever emerged in training. He said developers have “no right whatsoever” to bring a new species of conscious beings into the world — a red line aimed squarely at Anthropic’s model-welfare approach.
One ChatGPT Link Could Take Over Your Account: Inside the P1 Exploit Chain — and the VOLTA Malware That Uses AI to Price Its Victims
Security firm Parameter.ai disclosed a P1 ChatGPT account-takeover vulnerability: a crafted shared-conversation link bypassed the sandbox and CSP to run arbitrary JavaScript inside the victim’s signed-in session, reading private chats and exfiltrating data. OpenAI has patched it. The same week, Cisco Talos detailed the VOLTA infostealer, which uses LLMs to rank victims by monetization value — AI security is under attack from both sides.
Proton CEO Cancels AI ‘Re-Interviews’ After ‘Loser Thinking’ Leak: Inside the Rebellion at the Privacy Giant
Proton, the privacy company, is reeling after CEO Andy Yen’s ‘loser thinking’ Slack rant leaked and a company-wide AI ‘re-interview’ plan sparked an internal rebellion. One day after a 404 Media investigation, Yen cancelled the interviews — but AI metrics will still enter performance reviews, and users are cancelling subscriptions.
Mistral Large 4 Goes Fully Open-Weight: Trillion-Parameter Model Tops Open-Source Coding Charts in Europe’s Sovereign AI Bet
France’s Mistral has released the full weights of its flagship Mistral Large 4 (Le Chonk). After three extra weeks of reinforcement learning, the trillion-parameter multimodal model now tops open-weight coding benchmarks (FrontierCode 2.0 at 19.7%, Terminal-Bench 4.0 at 41%) and scores 51% on cyber defense, autonomously finding and patching vulnerabilities. Trained entirely in Europe and self-hostable, it anchors Mistral’s Sovereign AI strategy — though self-hosting demands ~1.06 TB of weights and serious compute, and its license restricts some hosted services.
Google’s Gemini 4 Carbon Tops AI Charts — But You Can’t Use It Yet
Google has released Gemini 4 Carbon, its latest frontier AI model, with major gains in coding, agentic tasks, and cybersecurity. It instantly tops the Artificial Analysis Intelligence Index, beating Claude Opus 5.5 and GPT-6 Astra, while leading both SEC-bench Pro and ExploitGym security benchmarks. Google says it can run for days on complex engineering tasks, halves unsupported claims, and starts at $2 per million input tokens — yet it’s currently restricted to select cybersecurity defense firms, with wider access coming via the Fairwind Program.
Google Genie 4 Arrives: Real-Time Steerable World Model Generates 2-Minute Realistic Worlds With Sound
Google DeepMind has launched Genie 4, its newest world model, in research preview. The model generates the most photorealistic and interactive worlds yet, adds real-time audio, and lets users type natural-language prompts mid-simulation to reshape the scene on the fly. With scene consistency stretched to two minutes and Gemini acting as the director, the world model race has entered its real-time era.
BlackBerry Reinvented: QNX Backlog Tops $1.2 Billion as AI Robotics Becomes Its New Growth Engine
BlackBerry, the former smartphone king, has completed one of tech’s most dramatic pivots: its QNX real-time operating system now holds a record $1.2 billion royalty backlog, up more than $250 million in just two quarters. CEO John Giamatteo calls AI robotics one of QNX’s fastest-growing businesses, and with deep Nvidia Jetson integration, BlackBerry is quietly becoming the infrastructure winner of the physical AI era — its shares have surged over 140% this year.
Zero AI Payoff Outside Tech: Apollo Economist Warns as $680B Buildout Meets Bond-Market Red Flags
Apollo’s chief economist Torsten Slok reports that despite unprecedented investment, AI has delivered “zero financial payoff” to U.S. companies outside Big Tech: S&P 493 profit margins and revenue per worker remain flat. Meanwhile, a $680 billion AI infrastructure buildout is tripping bond-market warning signals, and surging token costs keep squeezing corporate margins. The AI bubble debate has reached a critical moment.