June closes today and it has been genuinely wild. GitHub Copilot bills are exploding, Google is losing researchers like a leaky faucet, and Alphabet just completed the biggest equity raise in corporate history to fund compute. Here is what actually matters.
GitHub Just Ended Flat-Rate AI, and Developers Are Furious
For four years the deal was simple: pay $19 to $29 a month, use as much AI as you want. GitHub Copilot ended that deal on June 1.
Now one agentic session can run $30 to $40 in credits. Developers are posting screenshots of bills jumping from $29 to $750. One Reddit post that just said "Goodbye, Copilot" got shared thousands of times. The announcement thread on GitHub's own forum drew 900 downvotes.
This was always going to happen. Every AI company sold unlimited AI at prices that made no economic sense. The GPU compute behind a long agentic session costs real money, and $19 a month was never going to cover it for power users.
The real question is where developers go now. Claude Code does not meter tokens on Pro. Cursor has stayed more predictable. GitHub handed its competitors a gift with this rollout, and it is hard to see how they win back the power users they just alienated.
Google Is Having the Worst Month of Any AI Lab, Ever
Six senior researchers gone in five months. Gemini 3.5 Pro missed its June deadline, a model that Polymarket was pricing at 97% chance of not shipping. The stock dropped 7% in a single session after John Jumper, Nobel Prize winner and AlphaFold creator, walked out the door to Anthropic.
Sergey Brin sent an internal memo saying they need to urgently bridge the gap in agentic execution. When a co-founder is writing fire-alarm memos, things are serious.
The six researchers who left cover exactly what Google needs right now: reasoning, training architecture, science, coding tools, and pretraining. That is not a coincidence. Anthropic and OpenAI picked off the people building the foundation of the next Gemini, and they did it in a five-month window.
Google is still a massive, well-capitalized organization with excellent engineers. But talent losses like this take years to recover from, not months. The July Gemini 3.5 Pro launch is going to need to deliver something genuinely impressive to shift this narrative.
Alphabet Raised 85 Billion Dollars, and That Is the Rational Move
The same week Google's stock was getting hammered, Alphabet closed an $84.75 billion equity raise, the largest in corporate history for AI infrastructure. Berkshire Hathaway put in $10 billion of that.
The logic is pretty simple. Sundar Pichai said demand for their AI solutions is currently exceeding available compute supply. If you believe that is true, and the numbers suggest it is, you raise capital and build more compute. Capex guidance for 2026 is $180 to $190 billion, and Pichai said 2027 will be even higher.
Buffett is not known for making bets on hype. A $10 billion Alphabet investment at a discount to February highs, entered when sentiment was terrible because of talent departures, is a fundamentals call. You do not walk away from $174 billion in annual operating cash flow and a 2 billion user AI product over researcher turnover.
The talent problem is real. The balance sheet is also real. Both can be true at the same time. June 2026 is the month those two things are running on parallel tracks, and it is genuinely interesting to watch.