AI Unfiltered: Daily Briefing on Real Shifts—Models, Money & Movers

Alphabet just closed an $84.75 billion equity raise for AI infrastructure — the largest corporate AI financing in history. Berkshire Hathaway anchored it with ten billion dollars. That number should sound like dominance. Here's what it actually sounds like: desperation dressed in capital. In the same week Google secured that record raise, six senior AI researchers walked out the door. Four of them — DeepMind veterans — left in SIX DAYS. Anthropic and OpenAI picked them up before the ink dried. The market noticed. Alphabet's market cap dropped $269 billion. So the money is flowing in one direction, and the people who know how to USE it are flowing in the other.

That's not a funding story — that's a retention crisis at the company that invented the transformer. Step back. The capital environment right now is WILD. Private equity and venture capital poured $179.33 billion into AI unicorn rounds in Q1 2026 alone. Rounds worth at least a billion dollars each. Market Intelligence flagged concentration risk — too much money chasing too few companies. If you're holding a diversified AI portfolio, you might want to check how diversified it actually is. DeepSeek raised $7.4 billion in its first external round. Valuation: north of fifty billion.

8090 — a startup building a natural language software development platform — closed $135 million. Their pitch: developers write structured documents called Requirements and Blueprints, and the AI builds the application. No code, just intent. BMW launched a $300 million fund targeting agentic AI and physical robotics. X Square Robot in Shenzhen closed a Series C above $2.8 billion, backed by Xiaomi, Meituan, Alibaba, ByteDance. The automotive and robotics sectors are no longer watching AI from the sidelines — they're writing checks with nine zeros. Now — the model releases.

OpenAI shipped a limited preview of GPT-5.6 last week. Limited, because the Trump administration asked for regulatory oversight before a full rollout. The model has expanded context windows and better coding capabilities, positioning it against Anthropic's Mythos. But "limited preview" is the operative phrase. OpenAI wanted a broad launch; the government wanted guardrails. So we got a compromise that satisfies neither speed nor safety. Meanwhile, Anthropic's Claude is MOVING. Paying users and revenue up 75% since January. They just launched Claude Tag — a persistent AI agent that integrates with Slack, carries memory across sessions, and handles asynchronous collaboration.

Enterprise adoption is real. The growth curve is steep. Here's the pattern worth watching. The companies raising the most capital are not necessarily the ones shipping the fastest or retaining the best talent. Alphabet has $84.75 billion and a talent problem. Anthropic has momentum and market fit. OpenAI has regulatory friction and a cautious release schedule. DeepSeek has $7.4 billion and a valuation that assumes flawless execution. The AI landscape today is not about who has the most money. It's about who can deploy it before the people who know how to build leave for the next thing.

Alphabet just learned that lesson the expensive way.