Insights on AI Usage for an Agentic AI Future

Boomers

Boomers use AIs as consultants, engaging in frequent short sessions for business coaching, content creation, and companionship.

Boomers use AI just one to three times a week, for five to fifteen minutes at a time, to write and create content. They look to AI as a business coach, helping them create promotional content for their businesses. Boomers who participated in our survey shared that they rely on AI apps for research, writing, search, and content creation. Several respondents described creating AI friends, with custom appearances, for social interactions.

“I use it sometimes for writing, but mostly for creativity and brainstorming—getting ideas for things I’m working on.”—Boomer participant quotation

Gen X

Gen Xers are frequently power users who are deeply integrating AI as assistants into their daily life and work tasks.

Gen X is increasingly adopting AI as a versatile virtual assistant, streamlining a wide range of daily personal and professional tasks and effectively leveraging AI to enhance their productivity, as well as to offload routine redundancies. They are using AI multiple times a day for practical tasks such as controlling their smart-home environment and devices, including lighting, climate, and cameras. They’re the generation most likely to be power users.

“I really enjoy using AI, and it’s something I use on a daily basis as a friend, as a coworker, as someone to assist me in my work and personal daily tasks.”—Gen X participant quotation

Millennials

Millennials tend to use AIs as mavens, who are capable of providing expert sources of knowledge, personal recommenders, and collaborative planners.

Millennials are using AI for planning, typically spending 30 to 60 minutes per session. Key use cases for our respondents were meal planning and nutrition, with 56% of Millennials citing such AI use cases. Of those using AI to aid in their job searches, 62% of respondents were Millennials. Some respondents are interacting with AI to help them refine their professional communication style, using the AI’s outputs to achieve a new tone in written and verbal communications.

“I’ve asked AI to play the role of knowledge keeper. I’ll share memories, stories, and personal preferences, so they’re able to gain an idea of my likes and dislikes, and provide personal recommendations.”—Millennial participant quotation

Gen Z

Gen Z experiments with using AI as a mentor, advisor, or even a friend, often blending professional and personal roles in a single interaction.

Gen Z is navigating independence, from starting careers to building routines, and AI fills the gap by handling flexible roles to satisfy these diverse needs. Whether it’s as a therapist, executive assistant, or friend, we saw Gen Z experimenting with AI to fulfill a variety of roles for advice inputs. Research remains a primary use case, with users explicitly stating that they ask questions, seeking quick information, often as a replacement for a traditional search engine.

“I asked them to be a life coach, providing me with guidance on major decisions in my career, relationships, and personal life.”—Gen Z participant quotation

Trust Is the Gateway to the Adoption of Agentic AI

Across all segments, people are eager to use AI more, but trust—particularly around data privacy—remains a significant barrier to their usage of AI. A much lower percentage of users responded that they perceive data that they’ve shared with AI tools as private. Interestingly, users reporting both the lowest and highest AI usage, AI dabblers and power users, reported the highest rates of seeing AI as private, suggesting that privacy might be a factor keeping casual users from becoming power users.

Even among the most frequent users, most say that interacting with AI doesn’t feel natural. The data showed that just over 60% of power users and over 80% of AI dabblers report that AI interactions don’t feel natural. While the specific factors that are impacting this feeling of inauthenticity might include a lack of personalization or the use of a direct tone, more research will be necessary to assess the actual cause.

These are critical insights for UX designers, suggesting that trust will be as important as utility in an agentic future. Building trust requires more than privacy statements that are buried under a Settings menu. We must design experiences that are transparent about what’s happening with user data, consistent in their behavior, and capable of expressing empathy in a genuine way. In other words, the experiences that succeed will be those that make users feel not only understood but also respected.

Personalization Without the Settings Menu

While personalization has the potential to create richer AI interactions, most users have not yet taken the steps necessary to adjust their AI’s tone, name, or personality. Power users are significantly more likely to customize AIs across various dimensions, indicating their desire for highly personalized AI experiences. Among the various dimensions of personalization, tone of voice, name, and personality are the most personalized elements of GenAI experiences.

When we asked consumers about their preference for an AI’s tone of voice when interacting with GenAI platforms, across all user segments, most cited that they preferred a direct tone as their communication style. As users become more deeply engaged with AI, the preference for a direct tone decreases slightly, while the use of a polite tone sees a modest increase. The use of a personal tone remains relatively consistent across all age groups, but is still not the choice that the majority of consumers make. Today’s personalization often requires navigating through static menus and toggles, making it more of a chore than a delight.

In the future of agentic AI, personalization should emerge naturally—through conversational prompts, ambient cues, and adaptive onboarding that quickly learns and reflects a user’s preferences. The most effective personalization will feel less like configuration and more like collaboration, evolving with each user interaction.

Designing for the Agentic Leap

Agentic AI shifts the design challenge from creating tools to cultivating relationships. That shift demands a different mindset. AI experiences must be approachable enough for AI dabblers to explore without friction, yet deep enough to reward ongoing user engagement. They must be capable of adapting in tone, style, and emotional resonance to make interactions feel more natural and personal.

We must build for scalability from the very beginning—enabling AIs to grow in complexity as the user’s trust and ambitions grow. Takeaways about the impacts of Agentic AI design include the following:

• Design for AI dabblers and beyond. To build widespread adoption of agentic AI, future experiences must appeal to a wide range of consumers, not just power users. Create low-friction, exploratory user interfaces that welcome people with limited experience and teach them how best to derive value from agentic AI.
• Work overtime to earn users’ trust—from utility to emotions. Even GenAI power users said that using AI does not yet feel natural to them. Agentic AI experiences require trust, and an AI must earn trust over time through genuine connection and reliable relationship building. Don’t just optimize for utility. Explore AI that reflects and adapts to complex tones with a broad range of emotional expression, from casual to caring to commanding.
• Make personalization truly personal. Move beyond settings. Power users in particular are looking for personalized interactions with AI, and in the future of agentic AI experiences, we believe that this desire will only expand. Use ambient cues, cultural references such as celebrity voices, and optimized onboarding experiences to enable users to personalize AI experiences more naturally.
• Plan for use-case scalability. Design AI products with a general focus that can smoothly transition between a wide range of use cases—for example, career, creativity, care, and connection. Our respondents showed us that key use cases remain fluid rather than their becoming fixed. As users become more familiar and comfortable with AI, they give it ever more complex tasks to solve. A scalable experience, one that grows in depth and complexity at the pace of the user, enables broad adoption and customer retention.

As AI moves from reactive assistant to proactive collaborator, the experiences we design need to be more adaptive, more empathetic, and more contextually aware than anything we’ve built before. The winners in the agentic AI era won’t just complete tasks; they’ll forge a sense of connection.

The opportunity for product teams and UX designers is clear: our role is to guide AI beyond being merely useful to being a collaborator in meaningful ways. This is a future worth anticipating. 

To read more on this topic, read our recent research report: “Anticipating Agentic AI”

I would like to thank my team at Punchcut for contributing the excellent research and the insights I’ve shared in this column—particularly Jodi Burke, VP Insights; Jackie Antig, Associate Director, UX Research; and Lidia Juarez, Senior Design Researcher.