2 Unstoppable AI Stocks That Warren Buffett and Berkshire Hathaway Own

Berkshire Hathaway holds meaningful stakes in Alphabet (≈1.7% of its portfolio) and Amazon (≈0.7%), lending an institutional vote of confidence as both companies position for stronger performance in 2026. Alphabet, up roughly 60% in 2025, has accelerated in generative AI, is expanding potential revenue streams (including selling its TPU chips), and looks positioned for continued growth. Amazon’s underlying business showed revenue growth of 13% in Q3 and AWS — while only ~18% of revenue — provided ~66% of operating income and posted its fastest growth quarter in years, setting up a potential 2026 comeback after a muted stock performance in 2025. Investors should weigh ongoing AI-driven monetization and cloud earnings concentration when reassessing exposure to these large-cap tech names.

Analysis

Market structure: AWS (AMZN) and Google Cloud/TPU (GOOG/GOOGL) are primary beneficiaries — expect profit pool concentration to increase as AWS already supplies ~66% of Amazon operating income while being ~18% of revenue, implying modest revenue moves drive outsized EPS impact. Smaller cloud pure-plays and commodity GPU vendors face margin pressure if TPU adoption or AWS price/scale advantages accelerate; advertising incumbents benefit if Google monetizes generative-AI search without major cannibalization. Expect 2026 to be the inflection window for cloud/AI monetization; market-share swings of 2–5ppt in enterprise cloud are plausible over 12–24 months.

Risk assessment: Key tail risks are antitrust action against Alphabet (legal action or structural remedies within 6–24 months), supply-chain or GPU shortages that lift NVDA pricing (benefiting competitors in the near term), and enterprise AI adoption stalls that would delay revenue recognition into 2027+. Hidden dependencies include data-center capex cycles and OEM partnerships; margins are highly sensitive (a 100bp margin swing in AWS shifts Amazon EPS materially). Watch quarterly AWS/Cloud ARR and TPU commercial revenue as primary catalysts and 10‑K/FTC filings for regulatory shifts.

Trade implications: Direct plays — establish controlled long exposure to AMZN and GOOGL sized 1–2% each for multi-quarter re-rating; use 12–18 month LEAP calls (time to Jan 2027) to skew upside while limiting cash. Relative trades — long GOOGL vs short NVDA (size asymmetric, e.g., 1.0% long vs 0.7% short) to express TPU-induced GPU displacement risk; consider buy-writes or collars to finance LEAPs if volatility rises. Rotate 2–4% from cyclical retail/commodity semis into cloud/AI leaders over next 3 months as data points confirm AI monetization.

Contrarian angles: Consensus misses the operating-leverage effect in AWS and potential low-margin TPU OEM pricing that could temporarily compress GPU ASPs — a scenario that can widen AMZN/GOOGL outperformance by 30–50% vs peers in 12–18 months. Conversely, the market may be underpricing regulatory tail risk for GOOG (probability-weighted downside of 20–30% in structural scenarios); institutional endorsement (BRK.B stakes) is supportive but small (≤2% portfolio), not a guarantee. Historical parallel: cloud leaders re-rated after durable revenue/GM beats (AWS 2015–2018); here, validate with two sequential quarters before enlarging positions.

Key Decisions for Investors

• Establish a 1.5% portfolio long in AMZN today: 60% via shares, 40% via Jan 2027 LEAP calls (target 30–50% upside vs current), set an initial stop-loss at -25% on the equity tranche and add if AWS revenue growth exceeds 20% YoY for two consecutive quarters.
• Initiate a 1.0–1.5% buy-and-hold position in GOOGL with a 0.5% purchase of Jan 2027 LEAP calls as asymmetric upside; if Google reports TPU commercial revenue >$500M in any two consecutive quarters, scale GOOGL to 2.5% of portfolio within 30 days.
• Implement a relative-value pair: long GOOGL (1.0%) vs short NVDA (0.7%) using options (long LEAP calls on GOOGL, short-dated (90–120d) call spreads on NVDA) to express TPU-driven GPU displacement; cap max loss at 3% portfolio via position sizing and defined-option risk.
• Reduce semiconductor/retail cyclicals by 2% and reallocate into cloud/AI leaders or an actively managed cloud ETF within 3 months; revert only if NVDA posts >15% revenue beat two quarters in a row or CPI/interest shocks widen quickly.
• Set event-monitoring triggers: buy 6–12 month protective puts on GOOGL sized 0.5% if formal DOJ/FTC filings or legislative action occurs (probability-weighted impact threshold: >20% chance of structural remedy), and increase hedges if two consecutive quarters miss AI monetization metrics.