The AI Boom Is Entering a New Phase: 2 Stocks to Buy Now, According to Wall Street Analysts

Morgan Stanley sees the AI investment cycle broadening from semiconductor and infrastructure providers to software companies and AI adopters. Amazon’s trailing-12-month operating margin reached 12.1%, up more than 3 percentage points in two years, and Morgan Stanley expects 17% by 2028; Datadog’s Q2 revenue rose 36% to $1.1 billion, non-GAAP EPS climbed 41% to $0.65, and full-year revenue growth guidance was raised to 30%. The article highlights potential AI-driven growth but cautions that Datadog’s 104x adjusted-earnings valuation is expensive and risky.

Analysis

The investable distinction is between AI usage that produces measurable cost savings and AI exposure that only increases spending. For Amazon, automation could translate into durable retail operating leverage, but the market needs evidence that productivity gains outpace the associated labor, depreciation and fulfillment investment. The key validation is sustained retail margin improvement—not robotics deployment counts. Benefits may also pressure third-party logistics and labor-intensive retail competitors if Amazon converts efficiency into lower prices.

Datadog has a credible volume tailwind as AI workloads create more systems to monitor, but AI usage is not automatically incremental software revenue: customers can limit telemetry ingestion to control cloud bills, and cloud providers or observability rivals can compete for that spend. The crucial proof point is continued growth and guidance strength alongside healthy customer expansion, not product launches alone. At the valuation cited, any deceleration could overwhelm operating progress.

Near term, this is a narrative broadening catalyst, not evidence that semiconductor demand has peaked; avoid treating the software thesis as a reason to short NVIDIA or Broadcom. Over 1–3 months, earnings commentary on Amazon fulfillment productivity and Datadog customer usage/retention should discriminate winners. Over 6–18 months, realized cost savings could support Amazon earnings quality, while Datadog’s payoff depends on converting AI workload growth into durable net expansion. Contrarian risk: investors may be extrapolating AI adoption before companies disclose attributable savings or software monetization.

Key Decisions for Investors

• Favor a measured long in AMZN over adding to high-multiple AI software exposure: scale in around earnings or weakness, and reassess if retail operating margins stop improving or management signals productivity gains are being absorbed by fulfillment costs.
• Keep DDOG as a small, higher-volatility exposure rather than building aggressively at the valuation cited. Add only after a material pullback or evidence that customer expansion and revenue guidance remain strong; reduce if growth decelerates or usage economics weaken.
• Do not short NVDA or AVGO solely on the expected broadening of the AI trade. Their infrastructure demand may coexist with software adoption; revisit only if orders, guidance, or sector spending data show an actual deterioration.
• Track Amazon retail margin and fulfillment-cost disclosures, plus Datadog net retention/customer expansion and AI-related monetization where disclosed. Failure to show operating leverage or sustained customer spend would falsify the respective long thesis.