UBS Projects AI Spending to Reach 1.4 Trillion by 2027
UBS forecasts artificial intelligence capital expenditure will climb to 1.447 trillion dollars by 2027, driven primarily by surging memory costs.
UBS projects that artificial intelligence capital expenditure will reach 998 billion dollars in 2026 and climb to approximately 1.447 trillion dollars by 2027. This trajectory marks a steep increase from the 506 billion dollars recorded in 2025.
The financial services firm identifies surging memory costs as the primary driver of this growth. Memory spending is estimated to rise from 71 billion dollars in 2025 to 923 billion dollars by 2027, increasing memory's share of total AI capital expenditure from 14 percent to 64 percent over that period.
UBS notes that if these spending increases result from price hikes rather than volume, the capital will primarily function as a transfer of income and profits to memory producers in Asia. Such a shift would contribute more to the GDP of those Asian economies than to the real GDP of the United States.
