Good morning, India just pledged $210 billion to become an AI superpower, a Microsoft bug quietly fed your confidential work emails to its AI without permission, and Fei-Fei Li just raised $1 billion to teach AI to understand 3D space. Here’s what happened 👇
1. India Just Made the Biggest AI Bet in History
At India’s AI Impact Summit in New Delhi today, the numbers got staggering fast. Reliance — India’s largest company — committed $110 billion to AI infrastructure. Adani pledged another $100 billion. That’s $210 billion from just two companies, aimed at turning India into one of the world’s biggest AI hubs. Meanwhile, OpenAI signed its first major deal with Tata Group to build 100 megawatts of AI-ready data center capacity in India (with plans to scale to 1 gigawatt), and hundreds of thousands of Tata employees will get access to ChatGPT Enterprise. The event drew Sam Altman (OpenAI), Dario Amodei (Anthropic), Sundar Pichai (Google), and even Emmanuel Macron — though Bill Gates pulled out hours before his keynote, citing unspecified reasons.
The most candid moment of the day: when Prime Minister Modi asked all the executives on stage to raise their hands together in a symbolic show of unity, most obliged. Two didn’t — rival CEOs Sam Altman and Dario Amodei.
Why it matters: The AI race is no longer just a US-China story. India is writing $210 billion checks to get in the game. For everyday people, more AI infrastructure means faster, cheaper, and more localized AI services — especially for the 1.4 billion people who live there.
Source: Reuters | TechCrunch
2. A Microsoft Bug Was Feeding Your Confidential Emails to Its AI — For Weeks
Microsoft confirmed that a bug in its Copilot AI was silently reading and summarizing confidential emails inside Microsoft Office — even when companies had specifically set up policies to prevent that from happening. The bug affected Microsoft 365 customers using Copilot Chat, and it’s been happening since January. Emails marked as “confidential” were incorrectly processed by the AI, bypassing data loss prevention policies that organizations put in place to keep sensitive information out of AI systems. Microsoft says it started rolling out a fix earlier in February, but hasn’t said how many customers were affected.
Why it matters: You pay for software. You set up security policies. And the AI reads your confidential emails anyway — for weeks — without you knowing. This is exactly the kind of story that should make you think twice before pasting sensitive information into any AI tool, including the ones built into software you already use every day.
Source: TechCrunch
3. The Woman Behind ImageNet Just Raised $1 Billion to Teach AI About the Physical World
Fei-Fei Li — the Stanford professor who created ImageNet, the dataset that kicked off the modern AI era — has raised $1 billion for her startup World Labs. The biggest chunk, $200 million, came from Autodesk (the company behind AutoCAD, used by architects, engineers, and filmmakers everywhere). Other backers include AMD, Nvidia, and Fidelity. World Labs is building what’s called a “world model” — AI that doesn’t just process text or images, but actually understands 3D space, physics, and how the real world behaves. Their first product, Marble, lets users generate editable 3D environments from a text prompt. The Autodesk partnership starts with entertainment — think AI-generated 3D worlds for games and films — but the long-term vision is AI that can design buildings, simulate factories, and reason about physical systems.
Why it matters: Most AI today understands words and images. World Labs is betting the next frontier is AI that understands space — which is how humans actually experience reality. This has massive implications for architecture, manufacturing, filmmaking, and robotics.
Source: TechCrunch
4. Google Just Added an AI Music Maker to Gemini
Google’s Lyria 3 — its AI music generation model — is now rolling out inside the Gemini app. You can describe what you want (in text, or based on an image or video), and Gemini generates a 30-second music clip. It’s still in beta, and the results are described as “something like music” rather than studio-quality tracks. But this is Google’s most direct consumer push into AI-generated audio yet, following similar moves from OpenAI, ElevenLabs, and Suno.
Why it matters: AI music generation is moving from niche tools into the apps hundreds of millions of people already use. Whether you need a quick background track for a video or just want to play around with what’s possible, this is now one tap away in Gemini.
Source: The Verge
Quick Hits
-
Perplexity ditches AI ads: The search startup announced it’s abandoning plans to place ads in its AI results, with executives saying ads could have users “doubting everything.” A notable stance as ChatGPT moves in the opposite direction. (The Verge)
-
Netflix threatens ByteDance with immediate litigation over Seedance AI: Netflix gave ByteDance a 3-day deadline to stop its Seedance AI from generating content based on Stranger Things, Squid Game, Bridgerton, and other Netflix properties — calling it “a high-speed piracy engine.” (The Verge)
-
Meta is spending $65 million to influence AI legislation: The company is funding two new super PACs — one targeting Republicans, one targeting Democrats — to back politicians friendly to AI and fight regulation that could limit Meta’s AI business. (The Verge)
That’s it for today. The story of February 19, 2026 is really about one thing: who gets to control the infrastructure AI runs on. India is betting $210 billion it’ll be them. Microsoft’s bug is a reminder of what’s at stake when the infrastructure already inside your laptop goes wrong. And World Labs is asking a different question entirely — not just who controls AI, but whether AI can finally understand the world the way humans do.
Forward this to someone who needs to stay in the loop.
AI for Common Folks — Making AI understandable, one concept at a time.
Leave a Reply