September 8, 2026
Bonus Content: Google’s AI Search Is Growing. The Margin Question Isn’t Answered.
The Defense Story Most Investors Miss
A military arsenal gets the headlines. But a supply arsenal determines whether any of it can actually be built.
Before a missile, aircraft or infrared system reaches a production line, manufacturers need the critical materials inside it. One of them is antimony, and America produces virtually none from new mines.
That is why Washington is now trying to rebuild both arsenals at once. Pouring billions into weapons production does not solve the problem if America still depends on foreign sources for the materials inside them.
One emerging company has already assembled several antimony projects on friendly ground, led by a historic district now receiving a modern second look.
The major contractors will capture the headlines. But if their products depend on a material America barely produces, the quieter opportunity may sit much closer to the beginning of the chain.
Follow the company behind the defense trade most investors overlook >
Google’s AI Search Is Growing. The Margin Question Isn’t Answered.
GOOGL sits near $336 after enduring its longest monthly losing streak in over a decade through the summer. The September rebound is real, but the stock is still roughly 13% below its August peak of $384. The question traders need to answer is not whether Google is competing in AI. It clearly is. The question is whether the architecture of its monetization holds as reasoning models take over the search experience.
Start with what we know. Google Search & other revenue grew 17% year over year to $63.3 billion in Q2 2026, and Google has said AI Mode surpassed one billion monthly active users. That is not a company losing its search franchise. Just one year after its debut, Google says AI Mode has surpassed one billion monthly users, with queries more than doubling every quarter since launch. The volume story is intact.
The monetization story is where it gets complicated. Google says AI Mode queries run three times longer than traditional searches, and the company has argued that extended length is creating new ad inventory, with Gemini’s understanding of intent increasing its ability to deliver ads on longer, more complex searches that were previously hard to monetize. On paper, that is bullish. In practice, the per-query economics are still being assembled in real time.
Google says its AI Max tool is opening billions of previously unmonetized searches to ads, with the company’s chief business officer saying it helps match ads to searches that were too complex for traditional keyword targeting. AI Max is now out of beta with more than 500,000 advertisers, and Google says users of AI Max or Performance Max are seeing an average 15% lift in conversions or conversion value at a similar return on ad spend. That is the revenue bridge Google is building between the old keyword model and the new reasoning-first model.
But the uncertainty is not hidden. Alphabet has not separately disclosed AI Mode revenue, revenue per query, or advertising click-through rates, and user growth alone cannot prove that monetization is working. The question that actually matters has moved one layer deeper: whether the monetization model of AI-answered queries can sustain the margins that made Search the most profitable advertising product ever built, and that question will take several more quarters to answer properly.
Meanwhile, the capital commitment is not slowing. Alphabet’s 2026 capital expenditure guidance has been updated since the original $175 billion to $185 billion range. That is a bet on infrastructure that will pressure free cash flow before it pays off.
For traders, the structure here is asymmetric in both directions. The bull case is that strong usage traction across AI Overviews, AI Mode, and Gemini suggests Google is effectively transitioning search users to AI users, and that the monetization layers catch up to usage. The bear case is that longer, richer answers reduce click-through, that ad density stays light by design, and that the capex bill arrives before the revenue does. GOOGL at $336 is not pricing in either extreme. It is pricing in patience. Whether that patience is warranted depends on what Q3 earnings reveal about per-query revenue trends, and those numbers land in late October.
