Alphabet’s $85B AI War Chest Meets Washington’s 30-Day Safety Window

Alphabet raises $84.75B for AI expansion while navigating a new voluntary federal AI testing framework; record cloud revenue and $180B+ capex planned.

Alphabet finds itself caught between two enormous forces: a record-breaking $84.75 billion equity injection to fuel its artificial intelligence ambitions, and a new White House order that could slow down the very product launches those billions are meant to support. The tension crystallized this week as investors digested the sheer scale of the capital raise alongside a voluntary federal review framework for cutting-edge AI models.

The capital increase, the largest in Alphabet’s history, was originally set at $80 billion but swelled to $84.75 billion after overwhelming demand. The structure included $40 billion in at-the-market sales, a private placement with Berkshire Hathaway, and public offerings of Class A and Class C shares — the latter bumped from $15 billion to $18 billion. A mandatory convertible preferred stock tranche carries a 6.25% dividend. The share tranches are set to close June 4, with the depositary shares following a day later. Executives used the concurrent investor day to hammer home a stark message: demand for AI solutions outstrips available capacity “to an extent well beyond our current capabilities.”

That demand is reflected in the numbers. First-quarter 2026 revenue hit nearly $110 billion, up 22% year on year, marking the eleventh consecutive quarter of double-digit growth. Google Cloud alone delivered a record $20 billion in revenue, while its operating profit tripled to $7 billion. The backlog in the cloud unit nearly doubled to $462 billion. Over the trailing twelve months, Alphabet’s operating margin stood at 33%, and operating income has tripled in five years.

Yet the spending spree is historic — and risky. Capital expenditure is set to hit $180 billion to $190 billion in 2026, six times the 2022 level of $31 billion, with further increases planned for 2027. Funding comes from operating cash flow of $174 billion over the past four quarters, debt issuance, and the equity raise. About $30 billion of the ATM program is earmarked for tax liabilities tied to employee stock this year.

Enter the regulatory wildcard. On June 2, the White House signed an executive order creating a voluntary federal testing framework for powerful AI models. Developers, including Google, must grant government access to their most advanced models up to 30 days before releasing them to trusted partners. The order explicitly excludes mandatory licensing or pre-approval, and the 30-day window is shorter than much of the industry had feared. Reuters reported that Anthropic, OpenAI, and Google had already met with the U.S. administration in May. Investment banks have characterized the order as moderate — more clarity than a laissez-faire approach but far less restrictive than a licensing regime.

For Alphabet, the timing is delicate. The company is in the midst of an expansion phase, building out compute capacity, cloud infrastructure, and model deployment. A 30-day review period must become routine, or delays could ripple through cloud services and developer products. The stock, meanwhile, has lost momentum. Shares traded around €311 on Wednesday, down roughly 7% over the past week and about 11% below the 52-week high of €350.75 reached in May. The relative strength index sits at 39.8, indicating waning buying pressure. The dilution from the $85 billion capital increase adds near-term weight, though the board raised the quarterly dividend by 5% to $0.22 per share, with a record date of June 8 and payment on June 15.

Alphabet’s Gemini ecosystem now counts 900 million monthly users and 350 million paid subscribers. Enterprise adoption is accelerating, with a 40% quarter-on-quarter jump in paying users. Gemini 3.5 Flash has already launched, and the Pro version is due in June. The company’s ability to turn massive AI investment into revenue growth is clear. The question for investors is whether the new regulatory framework will introduce just enough friction to blunt that momentum — or whether the 30-day review becomes a manageable speed bump on the road to AI dominance.