Bank of America analysts say that internet-oriented growth stocks including DoorDash (NYSE:DASH) and Uber (NYSE:UBER) should be well placed to outperform when the present phase of the artificial intelligence cycle evolves.
The bank interprets recent stock behavior in the internet sector as indicative of an early AI-cycle stage. Conference discussions have emphasized strong demand for AI processing capacity and a rapid increase in AI application development. Those conversations, Bank of America said, reveal a market dynamic where many internet companies are beating expectations, yet growth and momentum-focused funds are rotating into semiconductors and hardware to capture the immediate upside tied to capacity expansions. That flow of capital has weighed on internet sector returns.
According to the bank, the current AI cycle phase is likely to persist while hardware and semiconductor capacity is ramping to meet demand. Once that supply backdrop improves, services and applications built on top of the hardware infrastructure could be in a better position to perform.
eCommerce and agentic shopping
Bank of America convened Katie Wilson of Serious Moonlight Consulting and Corey Thomas of AMZ Atlas to discuss the emergence of agentic shopping - AI-driven shopping agents that guide purchases. The bank noted that assembling structured data sets is a vital prerequisite for these agents to function effectively. It also observed that traffic driven by agentic experiences remains limited today, and widespread adoption is expected to be gradual. Early gains are likely to concentrate in commodity categories where consumers can incrementally build trust in agent recommendations.
The bank reiterated that Amazon's physical logistics network and rapid delivery capability remain a core competitive advantage. It suggested AI-related improvements in conversion rates could encourage higher marketplace advertising spend.
Travel discovery, OTAs and advertising
In the travel sector, Bank of America said AI will reshape how consumers discover options and make decisions, but it does not foresee immediate disruption to online travel agency (OTA) booking volumes. The bank expects Google and large language models to primarily monetize agentic experiences through advertising in the near term. In that scenario, OTAs could see benefits from agentic traffic sourced by Google.
Early-stage AI adoption and model specialization
Bank of America hosted Dmitry Shevelenko, Chief Business Officer at Perplexity, who characterized AI adoption as still very early. Shevelenko noted that many workflows remain underpenetrated and that most users currently apply AI to basic tasks such as search and drafting emails. He also observed that large language model capabilities are diverging, with models increasingly specializing by task, a development that could support a multi-model ecosystem.
Autonomous vehicles and regulatory limits
Ren Chen, Chief Strategy Officer at WeRide, told Bank of America that robotaxis are expected to supplement ride-hailing fleets and to expand to roughly 20% to 25% of total fleets over time. Chen said that regulatory frameworks, rather than technical capability, constitute the main constraint on rapid scale-up of autonomous vehicle deployments.
The bank's commentary spans the supply-side constraints in semiconductors and hardware, the gradual realization of AI-driven consumer features in eCommerce and travel, and the regulatory hurdles shaping autonomous mobility adoption.