Human assistants often spend their days taking notes, booking travel, following up on outstanding material from other sources, and reminding their bosses of deadlines. As depicted famously in The Devil Wears Prada films, unlucky employees even might find themselves tasked with completing a range of personal requests, such as picking up dry-cleaning, compiling holiday gift lists, or booking private events. Across these tasks, whether reasonable or excessive, the common element is that assistants must function like a quiet, logical organizational center that runs in the background, dealing with the clutter of everyday chores and deadlines, so their employers can focus more consistently and clearly on their critical, creative, or executive tasks.
Such work thus requires someone who is detail-oriented and able to execute rote, boring tasks consistently and at nearly any time of day. Details, rote tasks, constant availability, working in the background, and precise consistency—those are the exact features that existing and emerging iterations of agentic artificial intelligence (AI) promise to offer.
In retailing contexts for example, mass merchandisers like Target and Walmart have integrated AI systems into their chat features, to help shoppers sift through vast information at exponentially quicker rates. Rather than sending a human assistant to the store to grab a host gift for the dinner the executive is attending that night, users can task ChatGPT to connect to third-party e-commerce checkouts and purchase a flower arrangement directly from its interface, as well as have those flowers sent to the site of the dinner, just before the party starts. Google Shopping’s speech-enabled bot even can conduct simple conversations with local stores to confirm the availability of specific items.
Other iterations are prominent in tourism sectors. Leading online travel agencies, including Priceline, Expedia, and Kayak, have introduced tools to support automated reservations, and Google is actively working to join the competition through its search sites. The travel-oriented AI tools handle cross-site aggregation tasks efficiently and effectively, such that they are able to search for ticket prices or resort discounts across multiple platforms, while setting alerts for updates and registering changes in real time. Recent estimates indicate that four out of five travel companies plan to adopt some form of autonomous AI in coming years.
With regard to users’ reactions to and reliance on such digital assistants, existing findings seem promising. In retail sectors, almost half of respondents to one survey reported that they had used AI to facilitate their holiday shopping in a recent year. That percentage rose dramatically when focusing solely on millennial and Gen Z survey respondents. Opinions about AI-augmented travel searches appear somewhat more mixed though. One survey suggested that approximately one-third of customers feel comfortable using AI to make travel plans, and according to the industry news site Skift, nearly all consumers it contacted had found reliable information through an AI-optimized search. But a State of Travel 2025 report suggested that only one in eight users indicated they would trust AI to make independent decisions. Virtually no one agreed that they had sufficient confidence to allow AI agents to make or change their travel itineraries without any oversight.
Emerging enhancements and new agentic iterations of AI promise to execute and complete more multistep processes independently, such that they appear capable of performing long-term planning and decision-making, similar to the way a human assistant might. But just as bosses might worry about the capabilities of a new human assistant, users remain skeptical of the performance achieved by AI-enabled assistants, especially for critical decisions like travel. Over time, an effective assistant who demonstrates their reliability and capabilities can gain their bosses’ trust. Accordingly, we might predict that if AI agents continue to perform well, they also will continue to take over more tasks from busy users.
Discussion Questions
- This abstract presents AI assistants performing shopping and travel booking tasks. What other regular or daily tasks might consumers be interested in assigning to an assistant, and how effectively can current AI systems perform those tasks?
- Using your preferred AI system or chatbot, ask it to book a hypothetical trip for your next vacation, using whatever parameters you prefer. Review its recommendations. How well did it do? Did this experiment make you more or less likely to trust travel information gathered by an AI agent, without confirming its recommendations?
Sources: Gabe Castro-Root, “What Is Agentic A.I., and Would You Trust It to Book a Flight?,” The New York Times, November 25, 2025; Natallie Rocha and Kailyn Rhone, “A.I. Can Do More of Your Shopping This Holiday Season,” The New York Times, November 25, 2025.









You must be logged in to post a comment.