Anthropic $30B Google AI deal highlighting compute as the new oil in global AI competition

​​AI leaders unite against IP theft as Anthropic powers $30B Google surge and Microsoft launches three speech, voice, image models

OpenAI, Google, and Anthropic are sharing intelligence via the Frontier Model Forum to detect unauthorised model distillation by Chinese AI models. The move targets large-scale extraction patterns linked to firms like DeepSeek, now flagged as a financial and security risk.

Anthropic signed a multi-gigawatt TPU deal with Google and Broadcom, with capacity expected by 2027. Microsoft also released MAI models covering speech, voice, and image generation. Apple reported App Store Submissions Jump 84% as a result of vibe-coding tools. Here are important details you should know.

OpenAI, Google, Anthropic Join Forces Against AI Model Theft

OpenAI, Google, and Anthropic now share intelligence through the Frontier Model Forum to detect Chinese AI copying. DeepSeek allegedly distilled US model capabilities without authorization. Anthropic has already identified three Chinese labs extracting its models, calling the threat a national security risk in global AI competition.

Anthropic’s $30B AI Boom: Gigawatts from Google Deal!

Anthropic signed a new agreement with Google and Broadcom for multiple gigawatts of TPU capacity, coming online in 2027. The Anthropic Google deal expands its November 2025 commitment to invest $50 billion in US computing infrastructure. Anthropic’s run-rate revenue has now crossed $30 billion, up from $9 billion in late 2025.

Microsoft’s Triple AI Launch:Speech, Voice & Image Models

Microsoft released three new Copilot AI updates under its MAI series, now available in Foundry and MAI Playground. MAI-Transcribe-1 covers 25 languages at 2.5x batch transcription speed. MAI-Voice-1 generates 60 seconds of audio per second. The last generative AI model, MAI-Image-2, delivers 2x faster image generation for enterprise use.

This App Delivers VIP Tennis Perks in Seconds

A major North American tennis tournament replaced manual concierge workflows with a centralized digital platform managing transport, bookings, medical, and dining requests. The system automated 70% of requests, reduced wait times by 80%, and improved guest satisfaction from 40% to over 85% through real-time coordination and personalization.

Vibe Coding Drove 84% Surge in App Store Submissions

App Store submissions jumped 84% year-over-year in Q1 2026, reaching 235,800 apps. Sensor Tower links the surge to the vibe coding AI trend, with tools like Claude Code and ChatGPT Codex enabling no-code AI apps at scale. Apple now processes over 200,000 submissions weekly, averaging 1.5-day review times.

Claude Code subscribers to pay extra for OpenClaw usage

Anthropic now blocks third-party tools like OpenClaw from using Claude Code subscription limits, moving users to a separate Anthropic AI subscription fees pay-as-you-go model. The policy covers all third-party harnesses. Anthropic cites unsustainable usage patterns as the reason, while still offering full refunds to affected coding AI tools subscribers.

Anthropic’s Urgent Initiative To Secure World’s Critical Softwares

Anthropic launched Project Glasswing with 12 partners including AWS, Google, Microsoft, CISCO, and Apple to use Claude Mythos Preview for defensive cybersecurity. The model autonomously found zero-day vulnerabilities in every major OS and browser, including a 27-year-old OpenBSD flaw and a 16-year-old FFmpeg bug. Anthropic commits $100M in usage credits to the effort.

GPS + RFID SOS Hack: Logistics’ Wild Tech Upgrade

Well-configured GPS, RFID, and SOS integrations enable reliable event transport operations. Optimized update intervals, correct RFID placement, and monitored alerts improve real-time visibility, reduce response delays, and support faster dispatch decisions through structured go-live validation and integration readiness checks.

Meta Halts Work With AI Start-up Mercor After Data Breach

Mercor, a key data vendor for OpenAI, Anthropic, and Meta, suffered a security breach that may have exposed proprietary AI training datasets. Meta has indefinitely paused all work with the firm. OpenAI is investigating potential exposure of its own data. Breached training data can reveal model-building secrets directly to competitors, making Meta AI partnerships a competitive risk.

Meta Built AI Shopping Tools With One-Tap Checkout

Meta introduced one-tap checkout via PayPal and Stripe at ShopTalk 2026, with Adyen and Shopify coming soon. New Meta AI shopping features let creators tag products in Facebook posts and Instagram Reels for affiliate commissions. E-commerce AI integration now surfaces product reviews and recommendations directly within ads.

Google’s Free Offline Voice Dictation App: AI Edge Eloquent

Google launched Google AI Edge Eloquent, an iOS dictation app built on on-device edge computing voice models. It transcribes speech, removes filler words, and refines text locally, with optional Gemini-based cleanup. The app joins a growing category of free AI productivity tools focused on fast, accurate transcription.

Designing Priority Transport System For Every Group

Event planners are separating VIPs, staff, athletes, and accessibility riders into distinct transport systems. This reduces congestion, avoids timing conflicts, and improves service precision. Structured zone access, dedicated routes, and aligned vehicle types ensure each group operates efficiently without competing for shared resources.

Google Maps can now write captions for your photos using AI

Google introduced Gemini AI image description features in Google Maps to auto-generate captions from user photos and videos. The update simplifies content sharing, improves contribution tracking with points and badges, and strengthens its 500M+ contributor network supporting real-time place insights.

Claude Conway: Anthropic’s 24/7 AI Agent Awakens

Anthropic is testing Conway, a sidebar-based environment where Anthropic autonomous AI agents run continuously inside Claude. It adds Search, Chat, and System controls, supports webhooks and Chrome actions, and introduces installable extensions, pointing to persistent, tool-connected agent workflows.

What Else Is Happening?

Subscribe to our tech newsletter to receive regular insights that keep you informed. And if you find it valuable, please share it with your network to help spread the word. 
Catch you next time with fresh insights on AI and Tech, right here.