🧠 Arm unveils new AI chip, expects it to add billions in annual revenue ↗
Arm rolled out a new data center chip called the AGI CPU, and this feels bigger than a routine product launch - more like a strategic pivot. The company says the chip is built for agentic AI workloads, the kind where software acts with minimal oversight instead of simply answering prompts. (Reuters)
That matters because Arm is trying to stake a firmer claim in AI infrastructure, where the money is loud and the competition louder still. Investors seemed to like the pitch as well, with Arm saying the new line could add billions in annual revenue. A notably direct claim. (Reuters)
⚖️ EU antitrust chief meets Google, Meta, OpenAI, Amazon CEOs amidst AI scrutiny ↗
Europe is turning up the heat across the AI stack - not just chatbots, but training data, cloud infrastructure, the whole apparatus. Teresa Ribera met, or was set to meet, top leaders from Google, Meta, OpenAI, and Amazon as the EU weighs how dominant platforms might use AI to lock rivals out. (Reuters)
So yes, this is about competition law, but it is also about who gets to own the rails beneath AI. In an almost old-school way, the chatbot race is starting to resemble a utilities battle - just with foundation models instead of pipelines. (Reuters)
🛡️ Databricks bought two startups to underpin its new AI security product ↗
Databricks launched a new security product called Lakewatch and, tucked into that announcement, revealed it had bought Antimatter and SiftD.ai. Lakewatch handles classic SIEM work such as threat detection and investigation, but with AI agents powered by Anthropic’s Claude. (TechCrunch)
It is a neat signal of where enterprise AI is heading - less spectacle, more "please make the security stack less painful." Terms were not disclosed, which is entirely in character for acquisition news, but the larger point is clear enough: Databricks is using its giant war chest to turn AI into infrastructure rather than ornament. (TechCrunch)
🎬 OpenAI’s Sora was the creepiest app on your phone - now it's shutting down ↗
OpenAI is shutting down Sora, the TikTok-like social app built around AI video creation. The company gave no reason and did not say when the shutdown would fully take effect, which makes the whole thing feel a touch abrupt. (TechCrunch)
Behind the scenes, Reuters reported the move also disrupted a major Disney-linked partnership tied to Sora, as OpenAI shifts attention toward coding tools, enterprise products, robotics, and AGI work. So Sora is out, just like that, even though it once looked like one of the flashiest doors into consumer AI. (Reuters)
💸 With $3.5B in fresh capital, Kleiner Perkins is going all in on AI ↗
Kleiner Perkins raised $3.5 billion across two funds, with $1 billion for early-stage bets and $2.5 billion for later-stage growth companies. The firm’s latest haul is noticeably larger than its last major raise, and the AI angle is not subtle at all. (TechCrunch)
The logic is fairly straightforward - Kleiner already has exposure to fast-growing AI names like Together AI, Harvey, OpenEvidence, and Anthropic, so it is doubling down while the market still feels half-frenzied, half-obscured. A mild contradiction, perhaps, though that is venture right now: cautious swagger. (TechCrunch)
🤖 Agile Robots becomes the latest robotics company to partner with Google DeepMind ↗
Agile Robots struck a strategic research partnership with Google DeepMind to put Gemini Robotics foundation models into its machines. The plan is to test, fine-tune, and deploy those systems in industrial settings like electronics, automotive, data centers, and logistics. (TechCrunch)
This is another sign that "physical AI" is becoming the phrase everyone wants to claim, borrow, or quietly pursue. Agile says it has installed over 20,000 robotics solutions already, so this is not some lab-only moonbeam - it is AI trying to climb out of the chat window and into warehouses, which is slightly unnerving and quite significant. (TechCrunch)
FAQ
Why does Arm’s new chip matter for AI infrastructure?
Arm’s AGI CPU appears significant because it is being positioned as more than a routine server chip. The company is aiming at agentic AI workloads, where software takes actions with less human supervision. That gives Arm a clearer path to compete more deeply in AI infrastructure, rather than simply supplying designs at the margins. The revenue claim also suggests Arm sees this as a major business line, not a side experiment.
What does the EU’s AI scrutiny of Google, Meta, OpenAI, and Amazon actually mean?
These meetings suggest Europe is looking beyond chatbots and examining the underlying power structure of AI. That includes training data, cloud platforms, and whether large companies can use their position to shut out rivals. In practice, the issue is not just what models can do, but who controls the rails they run on. AI regulation is therefore becoming as much about market structure as it is about safety.
How is Databricks turning AI into a security product?
Databricks used Lakewatch to present a practical enterprise AI use case: easing the strain on security operations. The product takes on work such as threat detection and investigation, while relying on AI agents powered by Anthropic’s Claude. The startup acquisitions matter because they reinforce the product beneath the surface, not just the marketing around it. It is a sign that buyers want AI woven into core workflows.
Why is OpenAI shutting down Sora if AI video looked so promising?
The article presents Sora’s shutdown as abrupt because OpenAI did not provide a reason or a full timeline. At the same time, reporting points to a broader shift toward coding tools, enterprise products, robotics, and AGI work. That suggests flashy consumer AI video may no longer be the top internal priority. In many companies, attention shifts quickly toward areas with clearer strategic or commercial payoff.
What does Kleiner Perkins’ new funding say about where the AI market is heading?
The new funds show that investors still see AI as worth backing across both early-stage and later-stage companies. Kleiner’s existing exposure to names like Anthropic, Harvey, Together AI, and OpenEvidence makes this look like a conviction bet rather than casual trend-chasing. The article also hints at the market’s mood: energetic, competitive, and still somewhat uncertain. Capital is flowing all the same.
Why are robotics partnerships becoming such a big part of the AI story?
The Agile Robots and Google DeepMind partnership points to AI moving from software interfaces into physical systems. Rather than remaining inside chat windows, foundation models are being tested in industrial environments such as logistics, automotive, electronics, and data centers. That matters because deployment in the physical world raises the stakes on reliability and practical value. It also shows AI growth is no longer confined to models and apps.