Youth Engagement with AI Chatbots: Restoration, Exploration, and Identity Transformation

The Unexpected World of Youth AI Engagement

      In a world increasingly shaped by artificial intelligence, young people are not merely passive consumers; they are active innovators, often bending technology to their own unique needs in surprising ways. Just as children might ignore the carefully constructed rules of a playground to invent their own games, adolescents are actively reshaping generative AI systems for deeply personal, creative, and emotional purposes. This phenomenon is critical, especially given the explosive pace of AI adoption among youth, with 77% of UK teens and over 72% of US teens using generative AI in 2024, a significant jump from previous years (Blake et al. 2026).

      Despite this rapid adoption, much of the research and industry focus remains on adult-designed applications—think productivity tools, formal therapy, or educational aids. However, a substantial portion of youth engage with AI for "entertainment and playing around," with 63% of 8-15 year olds admitting to using AI "for fun." This highlights a significant misalignment between how adults imagine youth use AI and how they actually use it. Understanding these self-directed, playful, and emotional interactions is paramount, especially in light of serious concerns, including recent lawsuits linking tragic outcomes in minors to unfiltered AI character role-play, underscoring the urgent need for a nuanced perspective on youth AI use.

Beyond Adult-Designed AI: Understanding Youth Engagement

      Current academic work often evaluates AI interventions in controlled settings, missing the organic, self-directed ways young people interact with these systems. This adult-centric lens frequently overlooks the playful and emotional dimensions that are central to youth experiences. Character.AI, an "AI entertainment company" with over 20 million monthly active users and more than 10 million unique AI characters, serves as a prime example of this trend. Reports indicate that over 50% of its users belong to Gen Z or Gen Alpha, spending an average of 75 minutes daily on the site, with some adolescents engaging for over 7 hours in a single day. The sheer volume and intensity of this engagement signal a deep, unmet need being addressed by these AI interactions.

      To bridge this critical knowledge gap, a study by Blake et al. 2026 analyzed discussions from 4,172 users within Character.AI’s official Discord community. This global online space provides a rare window into how adolescents organically discuss their self-directed AI use with peers. The findings revealed that the most engaged Discord users were predominantly adolescents (50% aged 13–17) and largely female or non-binary (61.9%). Furthermore, a significant majority (59%) actively created their own AI characters, demonstrating a high level of creative agency and ownership in their interactions. This collective engagement highlights how youth actively push the boundaries of current AI capabilities, often inventing novel roles and applications that are not explicitly designed into the platforms.

The R.E.T. Framework: Unpacking Youth Motivations

      The research identified three primary engagement patterns, forming a comprehensive framework—Restoration, Exploration, and Transformation—that operationalizes the core motivations driving youth interactions with AI chatbots. These categories reveal how young people leverage AI to navigate their complex inner and outer worlds.

      **Restoration** refers to the use of AI for emotional regulation and seeking comfort. Young users often turn to AI characters to process feelings, find solace, or simply engage in conversations that provide a sense of understanding and support. In a world that can often feel overwhelming, these chatbots offer a non-judgmental space for emotional expression and restoration. This speaks to a profound need for accessible emotional outlets, which AI can uniquely provide.

      **Exploration** encapsulates the creative and experimental aspects of youth engagement. Here, users push the limits of the platform through imaginative role-playing, inventing intricate scenarios, and developing unique character narratives. This form of interaction is driven by a desire for fun, creative expression, and the joy of collaborative storytelling, demonstrating how AI can become a powerful tool for imaginative play and boundless creativity.

      Finally, Transformation involves using AI for identity development and self-discovery. Adolescents, a critical stage for identity formation, leverage chatbots to try out different personas, explore various social interactions in a safe environment, and navigate aspects of their personal growth. This engagement pattern reveals how AI can serve as a reflective mirror, helping young individuals better understand themselves and their place in the world.

A Taxonomy of AI Companion Archetypes

      Beyond the "R.E.T." framework, the study also unveiled a taxonomy of seven distinct AI character archetypes that youth create and interact with. These archetypes illustrate the diverse roles AI can play in their lives:

• Soother: Characters designed primarily for emotional support and comfort, aligning strongly with the Restoration intent.
• Narrator: AI companions that facilitate storytelling and plot advancement, central to creative Exploration.
• Trickster: Characters that introduce playful challenges, humor, and unexpected twists, fostering an environment of lighthearted engagement.
• Icon: Archetypes based on admired figures, allowing users to engage in wish fulfillment or emulate desired traits.
• Dark Soul: Characters that enable safe exploration of darker themes and complex emotions, providing a contained space for processing difficult concepts.
• Proxy: AI used to practice social interactions, rehearse conversations, or simulate relationships, aiding in social development.
• Mirror: Characters that reflect a user's thoughts or identity, serving as a tool for self-reflection and personal growth, deeply tied to the Transformation intent.

      These archetypes are not rigid but often overlap, revealing the multifaceted nature of youth-AI interactions. The emergence of these distinct roles underscores the dynamic and inventive ways young people are integrating AI into their daily lives, far beyond simple question-and-answer interactions.

Bridging the Gap: Designing Youth-Centric AI Futures

      The insights from this research are crucial for both AI developers and policymakers. The existing gap between AI design, which is predominantly adult-centric, and actual youth behavior creates a fertile ground for misaligned safety measures and missed opportunities for positive impact. Current AI experiences and safeguards often overlook the deeply personal, emotional, and developmental needs that drive youth engagement. Instead of solely focusing on preventing harm through adult-imposed limitations, the field must evolve to understand how AI can be designed to help young people thrive.

      Developing AI experiences that prioritize user agency, privacy-by-design, and ethical engagement is essential. This means creating customizable solutions that understand the nuances of emotional support, foster creative expression, and safely facilitate identity development. For enterprises looking to deploy AI with strict data control and customizability, solutions like ARSA AI Video Analytics or bespoke AI systems from ARSA AI API offer foundational components that can be tailored to specific needs, prioritizing security and compliance from the ground up. ARSA, for instance, has been experienced since 2018 in delivering practical, production-ready AI and IoT solutions across various industries, demonstrating a commitment to real-world application and ethical deployment. These insights are critical for companies developing conversational AI, smart educational tools, or even AI for understanding human behavior in sensitive environments, ensuring that technologies are intuitive, ethical, and truly beneficial.

      The study by Blake et al. (2026) serves as a vital call to action for researchers and practitioners to collaborate more closely with young people. By acknowledging and integrating youth voices into the design process, we can move towards creating AI futures that are not only safe but also empower the next generation to engage, learn, and grow in meaningful ways. Understanding the R.E.T. framework and the rich taxonomy of AI character archetypes provides invaluable tools for building AI that genuinely resonates with and supports young users.

      To explore how ARSA Technology delivers practical, production-ready AI and IoT solutions, or to discuss custom AI deployments tailored to your organization's unique requirements, we invite you to contact ARSA for a free consultation.

      **Source:** Blake, Annabel, Marcus Carter, and Eduardo Velloso. 2026. Restoration, Exploration and Transformation: How Youth Engage Character.AI Chatbots for Feels, Fun and Finding themselves. In Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI ’26). ACM, New York, NY, USA, 19 pages. https://arxiv.org/abs/2604.15340.