Alphabet’s AI Compute Allocation Dilemma Adds to Cash Burn Concerns

What happened

Alphabet faces a strategic quandary over how much AI compute to sell to cloud customers versus reserve for its own AI products, as flagged by a recent Business Insider report. This reinforces existing worries from the Q2 2026 cash burn of $5.9B and a capex run-rate near $175-185B, which already pressure free cash flow. The scarcity forces a trade-off: prioritizing Cloud revenue could starve internal Gemini development just as Gemini 3.5 Pro remains in testing, while hoarding compute for in-house AI may limit Cloud’s near-term growth and margin quality. The tension elevates the risk that Alphabet fails to efficiently convert its infrastructure buildout into durable returns, potentially eroding Cloud margins below the 28-30% threshold. Until management clarifies its allocation strategy, the WAIT rating on the stock stays intact.

Implication

The trade-off may force Alphabet to favor either Cloud revenue growth or proprietary AI advancement, delaying Gemini 3.5 Pro and risking Search monetization. If it leans toward selling compute, Cloud margins could slip below 30% as infrastructure-heavy revenue scales. Conversely, reserving compute for internal AI could cap Cloud’s near-term upside, clouding 2027 revenue visibility. This dilemma underscores the need for disciplined capex allocation beyond headline cash burn, making Q3 2026 free cash flow and margin data pivotal. Until Alphabet shows a balanced compute strategy with healthy returns, the stock is likely range-bound with a more attractive entry near $285.

Thesis delta

The new information highlights a compute allocation risk: selling compute to cloud customers may undermine Alphabet’s own AI product cadence and margin quality, while hoarding it could stifle Cloud revenue. This elevates the importance of management’s capex allocation discipline beyond simple cash burn, reinforcing the WAIT call until clearer evidence of balanced monetization emerges.