Berkshire Hathaway Just Bought More of This Brilliant Artificial Intelligence (AI) Stock

Berkshire Hathaway increased its Alphabet stake in Q2 by 48.1 million shares, deploying about $16.35B at roughly $340 per share—its biggest move in the quarter—signaling strong confidence in Alphabet’s AI outlook. The article cites Alphabet’s AI-led momentum: Google Search revenue rose 17% YoY and Google Cloud grew 82% YoY in Q2, alongside planned $200B of capital expenditures this year largely for data centers. Overall, the piece frames Alphabet as a cheaper-than-recent valuation entry point for investors, despite AI concerns about search disruption.

Analysis

The real signal here is not “Buffett likes AI,” it’s that one of the market’s most valuation-disciplined allocators is comfortable underwriting Alphabet’s cash flows even with a large AI capex bill. That matters because it lowers the probability the market will assign a permanent “search disruption” discount; over the next 6-18 months, the debate shifts from revenue at risk to ROIC on incremental data-center spend. If GCP keeps compounding while Search stays resilient, multiple expansion can continue without needing heroic assumptions.

Second-order, this is bullish for the AI infrastructure complex but not uniformly so. Alphabet’s spend supports demand for GPUs, networking, and power, which is constructive for NVDA on a 6-12 month lag, but it also raises the hurdle for cloud peers: if GCP is winning share while monetizing AI more efficiently, it can force AMZN and MSFT to defend with more capex, compressing free cash flow conversion in 2026. The market often rewards capex acceleration in the first phase and penalizes it later if incremental margins do not inflect.

The main risk is that the 13F read-through is stale and likely already embedded in a stock that re-rates quickly on every “AI is fine” data point. The consensus may be underestimating how much of the current thesis depends on one metric: sustained cloud margin expansion, not just top-line growth. If cloud growth decelerates into the 40s or operating margins slip while capex stays elevated, the stock can de-rate fast despite still-looking-strong headline growth.

Contrarian view: this is less a fresh alpha event than confirmation that the market’s most feared AI bear case is being monetized away. The trade is probably better on pullbacks than on chasing strength; the upside is durable, but the near-term catalyst path is weak unless the next earnings print shows AI spend translating into incrementally higher operating income rather than just bigger gross additions.

Key Decisions for Investors

• Add to GOOG/GOOGL on weakness rather than strength; preferred entry is a 3-5% pullback or a retest of the post-earnings support zone. Base case is a 6-12 month rerating as investors gain confidence that AI is additive to Search, not cannibalistic. Falsify if cloud growth slows sharply or capex rises without margin expansion.
• Pair trade: long GOOG/GOOGL vs short QQQ for a cleaner expression of idiosyncratic re-rating. This isolates the thesis that Alphabet’s AI spend is being underwritten by durable cash generation while broader mega-cap tech faces more multiple compression if rates back up.
• Use NVDA as a second-order beneficiary only on evidence that Alphabet’s capex remains above plan. Entry should wait for a better risk/reward after any semiconductor rally extension; the thesis breaks if hyperscaler capex guides flatten or if Google optimizes more of the stack in-house.
• Watch AMZN and MSFT cloud commentary over the next 1-2 earnings cycles. If GCP continues taking share, expect a defensive response in pricing and capex that could pressure free cash flow conversion across hyperscalers; that is a relative-short opportunity rather than an outright index short.
• No immediate trade on BRK.B solely from this filing; treat it as sentiment confirmation, not a catalyst. If GOOGL underperforms the market after the filing despite stable fundamentals, that would be the cleaner entry point than buying the news.