Rising AI adoption, improving monetization to drive the next leg of AI rally: UBS

UBS strategists anticipate U.S. AI adoption to exceed 10% by year-end, a significantly faster pace than prior technological shifts, indicating sustained growth for the sector. They advise investors to maintain AI as a core tech holding, recommending a portfolio rebalance by reducing exposure to the 'enabling' semiconductor layer in favor of 'intelligence' and 'application' software and internet companies, where greater monetization is expected. Additionally, UBS highlights substantial private market opportunities across venture capital, private equity, and real assets like data centers, advocating a diversified barbell strategy combining listed AI leaders with private market exposure.

Analysis

According to a UBS strategy note, the artificial intelligence rally is entering a new phase driven by accelerating enterprise adoption and improving monetization, providing continued tailwinds for the sector. U.S. AI adoption has surged from 5.7% in Q3 2024 to 9.2% in Q2 2025, with UBS forecasting it will exceed 10% by year-end—a milestone that took e-commerce 24 years to achieve. This suggests the theme remains in its early stages. Consequently, UBS recommends a strategic portfolio rebalance, shifting focus from the 'enabling' layer to the 'application' layer. Specifically, they advise reducing exposure to cyclical semiconductor companies from a 50-60% allocation to 40-50%, while increasing allocation to software and internet companies from 20-30% to 30-40%, as these are better positioned to capture value from monetization. The firm also highlights opportunities beyond public equities, advocating for a barbell strategy that includes private market assets such as venture capital for early-stage innovation and real assets like data centers for foundational infrastructure.

Key Decisions for Investors

• Investors should review their AI-related holdings, considering a tactical shift to reduce overweight positions in the 'enabling' layer, such as cyclical semiconductor stocks, and increase exposure to the 'application' layer, comprised of software and internet companies poised for stronger monetization.
• Given the analysis that AI adoption is still in its nascent stages despite rapid growth, it is advisable to maintain AI as a core long-term thematic allocation within a technology-focused portfolio.
• For those with the appropriate risk tolerance and capital, it may be prudent to explore a diversified 'barbell' strategy by complementing public AI leaders with private market exposure in venture capital, private equity, or real asset funds targeting AI infrastructure like data centers.