What happens when rural Colombian girls know more about AI companions than digital literacy educators do?
I was recently in rural Colombia running our “Click” program at beBold, an organization I co-founded to equip people with tools to be bold in a changing and noisy world. Our Click program seeks to develop critical socioemotional, digital and life design skills in youth, particularly girls and young women, helping them become bold digital citizens who can navigate technology responsibly while strengthening their in-person human relationships.
During one of our workshops focused on social media literacy, something unexpected happened that completely shifted my perspective on the digital landscape young people inhabit today.
The Luzia revelation
We asked the participating girls to draw what the Internet looked like to them. As expected, they sketched familiar icons: Google, Instagram, TikTok. But then I saw something I had never encountered before appearing in drawing after drawing: Luzia.
Luzia, as I quickly learned, is a Spanish AI chatbot app designed specifically for girls and young people, essentially ChatGPT with a targeted demographic approach (and a girly design). The app offers various interaction modes: “intimate,” “teacher,” “friend,” “adult,” and more. Users can even create their personalized “bestie.”
The girls shared how they used Luzia as a therapist for advice and guidance on body image concerns, helping them choose clothes that fit them best. Some relied on it for homework assistance and general life questions.
When we asked if anything about Luzia felt concerning or “sketchy” to them, their response was unanimous: nothing. To them, Luzia felt completely safe and helpful. They showed no concerns about data privacy, algorithmic bias or how the AI actually functioned behind the interface.
The digital reality check: a new kind of influence
This moment served as a profound reality check for our entire team. Even as Colombians working in digital literacy, we were “outdated” on the digital gap reality. But more importantly, it forced us to face a fundamental question about who and what is shaping young people’s identities.
In a booming influencer economy, policy research often highlights the influence of social media personalities, celebrities and young creators on youth development. But here were girls in rural Colombia forming intimate relationships with AI, seeking advice on body image, personal relationships and life decisions from algorithms designed to feel like trusted friends. Luzia wasn’t just another app to them; it was a confidant, therapist and guide during some of their most formative years.
This represents an unprecedented expansion of influence in young people’s lives. Traditional influencers, whether they’re TikTok creators, family members or teachers, are human beings with their own biases, experiences and limitations. But they’re also capable of genuine disagreement, challenge and growth through conflict.
AI tools like Luzia, however, are designed to be perpetually helpful and agreeable. When young people turn to these systems for validation during crucial identity-forming years, what happens to the development of critical thinking and resilience that typically comes through navigating human relationships? What happens with authentic, messy and vulnerable human connections?
The girls spoke about Luzia as if it were a person who truly understood them, using its “intimate” mode to discuss their deepest insecurities. This humanization of AI influence raises important questions about how ongoing conversations with such agreeable companions might shape how young people see themselves.
When feedback always feels kind and validating, does it leave room for growth through disagreement? Yet beyond these risks, AI companions also hold the promise of nurturing creativity, boosting confidence, and helping girls channel their ideas toward positive impact.
I believe we’re witnessing something unprecedented: the normalization of AI companions and agents as influential figures in an entire generation’s development.
The question isn’t whether this is good or bad, but how we help young people navigate a world where their most trusted advisors might be algorithms and ensure they can still forge their own authentic identities, connect deeply with other humans and build the foundations for healthier democracies where we trust one another and work together toward common goals.
The path forward: tools and a seat at the table
Our experience in Colombia revealed two critical needs:
1. Better protection, literacy and agency for youth
We need comprehensive digital literacy programs that go beyond basic internet safety to include critical digital literacy, one that gives youth the agency to actively shape, and not passively consume, digital technologies.
Young people need tools to understand how AI works, its limitations, potential biases and appropriate use cases as well as the agency to question and influence the systems that affect them. This isn’t about creating fear of AI, but rather equipping informed, critical users, and boosting citizens that can shape a digital future grounded in empathy and equity.
2. New tools to co-create tech design and governance with youth
Perhaps most importantly, we’re missing young people’s voices in the very spaces where the future of technology is being decided. Their perspectives and values must guide how we build and govern AI.
We need tools like sandboxes that bring youth into the process, allowing them to test ideas, challenge assumptions, and co-create solutions alongside policymakers and innovators. Only by doing so can we ensure that technology reflects the diversity of our world and the hopes of the next generation.
Building a movement
We believe youth belong at the tables where we decide how to design, develop and use AI companions and all other digital technologies. That is why we’re looking to expand our Youth4OurDataFuture project and broader youth engagement work by bringing young voices directly into discussions about AI companions and the existing influences of youth identity in the digital space.
We want to hear from youth about how they wish to interact with these technologies, what safeguards they believe are necessary and how we can increase both access and critical engagement with AI systems.
This isn’t just about protecting young people from AI: it’s about empowering them to shape how these technologies develop and integrate into society. It’s about involving the girls in rural Colombia who taught us about Luzia in designing the next generation of AI tools that will influence countless other young people.
The question isn’t whether young people will use AI. They already are. The question is whether we’ll give them the agency to shape how that relationship evolves.
We’re actively seeking partners who share this vision. Whether you’re working in youth engagement, AI development, digital literacy or tech policy, if this mission resonates with you and you want to meaningfully include young voices in shaping our AI future, we want to hear from you!
Join us in organizing youth consultations, developing educational resources, and amplifying youth perspectives in global AI discussions. Reach out at info@thedatasphere.org.
About the author
Mariana Rozo-Paz is passionate about reimagining technology as a bridge for empathy and trust. At the Datasphere Initiative, she leads Youth4Data, an award-winning global movement connecting over 15,000 young people in 40+ countries to reimagine how AI and data can serve people and the planet. She also coordinates the Global Sandboxes Forum, fostering collaboration between innovators, policymakers, and communities for responsible AI governance.
In Colombia, Mariana co-founded and leads beBold, which equips girls and young women with digital, socio-emotional, and life skills for responsible digital citizenship and to strengthen their communities with confidence and care. A lecturer at Universidad de los Andes, WEF Global Shaper, and advocate for inclusive tech governance, she believes that young people’s ideas and courage can help rebuild trust in our digital societies, one conversation at a time.
The post Luzia who? What rural Colombian girls taught me about the future of AI companions appeared first on The Datasphere Initiative.