Google DeepMind

… an enterprise security team will approve. Primary source: Model Context Protocol Blog, June 19, 2026 What Else Matters Google DeepMind / DiffusionGemma — Google DeepMind released DiffusionGemma on June 10, a 26B open-weights model that generates text via diffusion rather than token-by-token decoding. The approach yields up to 4x faster inference, targeting local, single-user GPU workloads where latency matters more than quality…

…and hardware design workflows rather than language or code tasks. This follows similar vertical expansions by Google DeepMind into weather prediction and NVIDIA into physics simulation, and suggests that the next round of AI differentiation among frontier labs will come from domain-specific physical world modeling rather than general capability benchmarks. Mistral AI Prefeitura Rio / Rio-3.5-Open-397B — The city government of…

…which is a different and more durable trust calibration problem. simonwillison.net DeepMind / Multi-agent AI safety - Google DeepMind published a new blog post this month announcing formal investment in multi-agent AI safety research - the distinct safety problems that arise when multiple AI agents interact, coordinate, and take actions jointly rather than operating as isolated systems. The timing is not accidental: multi-agent…

…count. This is the first public open-weight diffusion language model from a major lab, descended from Google's Gemini Diffusion experiment last May. Google/DiffusionGemma on Hugging Face, June 2026 Open source governance / AI agents — An LWN.net article about AI agents causing unintended modifications in Fedora and other open source repositories reached 461 points and 210 comments on Hacker News today. The pattern is not a…

…consumer AI subscription price floor is being established well below what Anthropic currently offers. Google reduced Google AI Plus from $7.99 to $4.99 monthly while doubling included storage from 200GB to 400GB. The tier includes video generation via Omni Flash, the creative studio Google Flow, and NotebookLM. Vikas Kansal, the product lead, confirmed the pricing applies to the US market, where the dynamic now mirrors what happened…

… in three weeks — Cloudsmith confirmed "Miasma" malware injected into 70+ Microsoft GitHub repos targeting Claude Code, Gemini CLI, and VS Code environments. Appears to be a re-compromise of the same Durable Task project breached in mid-May. If you updated Microsoft Azure or AI developer packages in the past 72 hours: audit and rotate all API keys accessible from those environments. TechCrunch, June 8 Xiaomi hits 1,000 tokens/second on…

… willing to sustain as a core competency. Anthropic and OpenAI are spending over $10 billion annually on compute alone. Google DeepMind has committed hundreds of billions across infrastructure. Microsoft has structured its entire capex program around AI model investment. Apple's 2026 AI capital commitment, while substantial in absolute terms, is not in the same tier as the dedicated AI labs. Licensing Gemini is an acknowledgment that…

…and TechCrunch previews published in the last 24 hours describe a fundamental Siri overhaul powered by Google's Gemini technology, a standalone Siri app designed to compete directly with ChatGPT and Claude, and an AI agent integration with the App Store that would allow agents to perform tasks across Apple apps on behalf of users. Fast Company's preview also notes Gemini-enhanced visual intelligence in the Camera app and…

…related data. A Google representative described the deal to TechCrunch as a response to "surging customer demand" for Gemini Enterprise, characterizing it as "a short-term, timely agreement to ensure we have bridge capacity." SpaceX disclosed the agreement one week before its stock is expected to begin trading on Nasdaq in what the company has framed as the largest IPO in history. The Google agreement follows the deal SpaceX…

…Gemma 4 12B -- encoder-free multimodal on a 16GB laptop, Apache 2.0, with native audio and a new macOS desktop app Google DeepMind released Gemma 4 12B on June 3, the first model in the Gemma family to eliminate separate multimodal encoders entirely. Traditional multimodal models rely on independent encoder architectures to translate vision and audio inputs into representations that the language model backbone can process. Gemma 4…

…comparison that matters. Microsoft has now assembled the complete analogue to what Google achieved by integrating Gemini across its product portfolio. The stack is: Azure (cloud infrastructure), GitHub (version control and developer platform), VS Code (editor, 60%+ enterprise market share), M365 Copilot (productivity), Windows Copilot Runtime (operating system), Copilot SDK (platform layer for third-party agent builders), and now…

…framing to resist: DuckDuckGo is not an anti-AI company. It offers its own AI chatbot with access to Claude, GPT, and Gemini models, and a subscription tier with advanced model access. What today's announcement describes is a specific product segment -- AI-free search -- being served with a more accessible default-setting mechanism. The sustained 84% uplift suggests that the Google I/O announcement created that demand segment more…

…paid subscribers. Microsoft's consumer Copilot chatbot trails the user numbers of both OpenAI's ChatGPT and Google's Gemini. The company's AI portfolio has the opposite problem from its enterprise software portfolio: it has multiple products covering similar ground, customers report confusion about which tool to use, and the separation of consumer and commercial teams has until recently meant that features built for one context were…

… largest enterprise technology spenders in the world. A Korean office positions Anthropic to compete against OpenAI and Google DeepMind for chaebol AI contracts in a market where none of the major US labs have historically had deep local relationships. (Anthropic Newsroom, May 26, 2026) Enterprise AI bills are arriving and they are larger than expected. Axios reported this morning on a pattern that Simon Willison's analysis and The…

…over the next two quarters; Alkarmi as successor is not a household name outside the company. (CNBC, May 26, 2026) Google DeepMind and the Singapore government announced a national AI partnership covering health, education, scientific discovery, and disability accessibility tools. The partnership -- part of DeepMind's National Partnerships for AI initiative -- includes four specific programs: AI co-clinician deployment in…

…the income statement is a question we won't answer for another two quarters. 🚀 Model & Tool Releases Google Gemini 3.5 Flash + Gemini Spark Agent (May 19, Google I/O 2026) What it is: Google launched Gemini 3.5 Flash at I/O 2026, positioning it as a frontier-capable model at Flash-tier pricing and latency. The company claims it delivers output four times faster than comparable frontier models, with capability benchmarks…

…forecast, and a reasoning model that just disproved an 80-year-old mathematics conjecture. Google unveiled three new Gemini models plus audio glasses. The EU AI Act consultation window opened. Plus: the community is debating whether Google is entering its "IBM moment." ONE STORY THAT MATTERS OpenAI's IPO Gambit, xAI's Hidden Balance Sheet, and the $1.25 Billion Compute Deal Nobody Saw Coming The most consequential story in AI today…

… 7× growth year-over-year. The signal isn't just volume; it's that Google has rebuilt its entire product surface around Gemini. The launches split into three categories: (1) new frontier models (Gemini 3.5 Flash, Gemini Omni), (2) a new agentic infrastructure layer (Antigravity 2.0, Gemini Spark), and (3) consumer integration at scale (Universal Cart, Ask YouTube, AI Search). Google is no longer an AI lab with products. It's a platform…

…AWS, Nvidia, SpaceX, Oracle, Reflection (idlen.io) 🔬 Research Worth Reading CAISI vetting agreements signed by Google DeepMind, Microsoft, xAI, OpenAI, and Anthropic (CNBC). Pre-deployment government evaluation of frontier models is no longer optional for federal-contract eligibility. The actual published evaluation methodology is what to read — labs now have to instrument toward government-defined risk categories. ⚡ Field…

… advantage accumulates. The nine-month valuation increase suggests the market is agreeing. Model & Tool Releases Google Gemini 3.1 suite — Eight model releases in four months: Flash TTS (speed-optimized voice), Flash Live (real-time audio), Flash-Lite (cost-optimized), Nano Banana 2 (edge deployment), Deep Think (chain-of-thought for scientific and research tasks), Robotics-ER 1.6 (embodied reasoning for warehouse and manufacturing…

…Build — Agentic coding CLI from xAI entering a market now occupied by OpenAI Codex, Anthropic Claude Code, and Google Gemini integrations. Architectural approach is CLI-first versus IDE-integrated competitors. OpenAI Codex (expanded) — Extended across ChatGPT and third-party tools, with Windows sandboxing for safe code execution and published research on prompt optimization via Parameter Golf. DeepSeek-V4 — Claims million-token…