Consumer agents may power the next phase of the AI trade

Podcast: Signal over Noise| Takeaways from Silicon Valley, on Apple and Spotify (6 mins)
Video: Mark Haefele on why the equity rally can continue (3 mins)
Video:Market Playbook | Why higher rates alone don't mean lower equities (6 mins)

Tech shares rallied on Monday after media reports highlighted the rapid early adoption of Meta’s personal AI agent Muse. The Philadelphia Semiconductor Index rose 4.3%, and the optimism extended to Asia’s AI supply chain on Tuesday. Equity benchmarks in Taiwan and South Korea both closed higher.

Launched on 8 September, Muse was the most downloaded free app on the US Apple iOS and Google Play stores as of Monday, recording more than 2.5 million downloads in its first 13 days. Over an equivalent period following its mobile launch, Apptopia estimated 1.8 million iOS downloads for Muse in the US and Canada, versus 1.3 million for ChatGPT. Whether its leadership can last remains to be seen, with several AI labs all reportedly working to launch competing offerings soon.

Without taking any single-name views, Meta’s Muse provides early evidence that consumer AI agents could achieve broad adoption. This not only offers new AI monetization opportunities beyond enterprise application, but also underpins growing demand for computing power and AI infrastructure. We discuss the implications for the AI trade.

Consumer AI agents could introduce a new revenue stream through transaction economics. Consumer AI monetization has so far centered largely on subscriptions and advertising. Agents could add a third channel by executing transactions, such as purchasing goods, booking travel, paying bills, or managing recurring services on users’ behalf. Platforms that facilitate these activities may be able to capture a share of transaction value, while task execution could also increase user engagement, retention, and willingness to pay. The opportunity remains at an early stage, but it could broaden the consumer AI revenue model from monetizing attention and access to monetizing actions.

The potential pathway to monetization should improve hyperscalers’ returns on AI investments. Wider adoption of consumer agents could generate revenue through transactions, subscriptions, and advertising, while increasing cloud utilization and engagement across platform ecosystems. This would provide a clearer link between infrastructure spending and revenue growth, helping monetization catch up with the substantial investment already made in computing capacity. Although we note that leadership at the application layer is likely to remain fluid. With new agents in the pipeline, platforms compete over distribution, proprietary data, commerce access, and customer relationships.

Higher demand for computing power should further underpin semiconductor and hardware leaders. Unlike conventional chatbots, AI agents can browse, retrieve information, plan, and execute multistep tasks, creating greater demand for hosted computing and inference capacity that is more CPU intensive. If consumer adoption strengthens, this should support demand across processors, memory, advanced packaging, substrates, servers, networking, power systems, and cooling. Rising capacity requirements should also support investment in semiconductor manufacturing equipment and data-center infrastructure. Performance will not be uniform, however, making diversified exposure to high-quality enabling technologies preferable to treating semiconductors as a single trade.

So, we maintain our constructive outlook on the AI trade, supported by rising adoption and monetization, as well as growing capital spending. We favor diversified exposure across the value chain, including high-quality semi and hardware leaders that offer strong earnings visibility and attractive returns on capital, as well as the largest AI spenders whose scale, diversified earnings streams, and growing AI monetization opportunities should provide resilience across a range of outcomes. Select infrastructure software and defensive tech companies should also offer opportunities.