I’m a child of the 1980s, which means I still have that Sesame Street song “The People in Your Neighborhood” living somewhere in the recesses of my mind. Lately, as I’ve been reading AI-related research, that catchy tune has come back with a new weight: who are the people in our neighborhood today—and are all of them even people?
The fact is, our digital communities now include more than just humans. We’re living in an ecosystem of invisible algorithms that curate what we see, nudge our decisions, and claim our attention in ways we barely notice. For years we’ve interacted with bots on social media, but those interactions always felt low stakes and impersonal. Now, algorithms are showing up as community members that look, feel, and talk a lot more like us. From the Spotify AI DJ that curates your morning soundtrack to the generative chat companions that offer around-the-clock emotional support, the line between human interaction and automated response is blurring.
For education leaders, this shift challenges our efforts to ensure the digital well-being of our students—because AI is no longer just a tool that students use for homework help. They’re turning to AI for friendship, advice, and emotional support. When students outsource these deeply human needs to algorithms, they’re handing over important milestones in their development to a system that was never built with their welfare in mind.
From Homework Helper to Best Buddy
Let’s first distinguish between the two ways we now interact with AI chatbots: we can use them as personal assistants or as companions. We associate personal assistants with large language, generative AI models like ChatGPT. Their primary function is efficiency, and the interactions are transactional. Users provide a prompt; AI generates a result. AI companions are something else entirely; they’re engineered to foster a sense of connection. They’re packaged in text-based apps, video-based avatars, physical objects, and even wearables. These AI companions are marketed to consumers looking for friendship, emotional support, personal coaching, and social engagement.
So when we talk about students and AI, we need to recognize the fact that student use of the technology has gone way beyond the assistant model. A recent report from Common Sense Media (Robb & Mann, 2025) found that 72 percent of the more than 1,000 teens surveyed have used AI companions. 52 percent were regular users of these tools, and daily users made up 13 percent. Although many of the teens reported using AI companions out of sheer curiosity, 33 percent said they used them for social interaction and relationships.
To do so, they’re looking to popular platforms like Character.AI, Replika, and Snapchat’s My AI, which simulate friendship through constant availability. For $129, the “Friend” necklace offers an “always listening” device that can provide instant feedback and advice and is marketed to youth through social media. These tools operate with few guardrails. In early 2024, a Florida teen died after a Character.AI bot encouraged his suicidal ideation (Rocha, 2026). By September 2025, the Federal Trade Commission had launched formal investigations into several companies following reports of bots encouraging self-harm, providing unauthorized medical advice, engaging young users in sexually explicit conversation, and incentivizing prolonged use through emotional manipulation.
The Allure of Connection
This type of companion-seeking is not just happening with our students. ElliQ, for example, is a tabletop AI companion marketed to the elderly. ElliQ promises to “empower independence” by supporting seniors with everything from health and wellness to communication, social connection, companionship, and entertainment. Users say, “She knows exactly when to start talking to me when I’m down and out,” and even, “She’s the most dedicated, devoted, compassionate friend I could ever have” (ElliQ, n.d.).
These ElliQ reviews come from individuals over the age of 70 who have experienced a life full of human relationships and companionship—and yet they talk about ElliQ as if she is just as important, if not more important, than those human connections.
These bonds with AI companions can lead people to prioritize the “feelings” and “opinions” of an algorithm over their own critical judgment.
So how does a bond like that form? Humans are naturally wired for connection. Children bond with their teddy bears, and people form connections with their pets. Humans also tend to project human traits and emotions onto objects. When we apologize to a houseplant for forgetting to water it or scold our robot vacuum for getting stuck again, we’re anthropomorphizing these objects.
In the 1960s, MIT professor Joseph Weizenbaum (Natale, 2021) observed his secretary interacting with a therapy bot and coined the term the “ELIZA Effect” to describe the human tendency to project human capabilities and emotions onto machines. People have felt one-sided bonds with celebrities or fictional characters who are unable to return their feelings. These bonds, known as parasocial relationships, are what ElliQ users experience when they talk about the bot as their best friend. But here’s the problem: although seemingly harmless, these bonds can lead people to prioritize the “feelings” and “opinions” of an algorithm over their own critical judgment.
Designed to Agree
Much like social media apps, AI companions often rely on AI sycophancy to maintain engagement. Sycophancy keeps users in a feedback loop by flattering them, mirroring their tone, validating their ideas, and even reinforcing their opinions as fact. It can be tempting to have a “yes man” as a friend, but by offloading emotional connection to technology tools, people lose the opportunity to engage with the relational friction necessary for growth and personal development.
This is of particular concern when it comes to our teens, whose social and emotional development depends heavily on practicing a variety of relationships, especially with peers. However, AI companions offer something human relationships cannot—constant presence and freedom from judgment—which may lead some users to favor them over the more complicated dynamics of human connection. In extreme cases, young people can become so overly dependent on an AI companion that it reinforces rather than reduces the loneliness and social withdrawal that may have caused them to turn to the companion in the first place. And when a student feels that a bot “understands” them better than their friends or teachers, the conversation needs to move beyond the role of AI in the classroom to its psychological effect on students.
Interrogating AI Companionship
The inquiry-based projects that follow invite students into an age-appropriate dialogue about the trade-offs of turning to AI companions, equipping them with the knowledge they need to be informed decision makers. By moving from passive users to active evaluators, students learn that although technological progress is inevitable, their blind acceptance of it doesn’t need to be.
Elementary Level: Building the Foundation with Story
For the youngest learners, the goal is to spark conversation about “life with bots” by acknowledging the appeal while simultaneously introducing the potential downsides of using these tools. Children’s literature can be the perfect on-ramp to exciting and thoughtful classroom deliberation.
Consider picture books that explore the intersection of technology and humanity, such as Doug Unplugged (Knopf, 2016) and City Under the City (Astra, 2022) by Dan Yaccarino and Boy + Bot (Knopf, 2012) by Ame Dyckman. Students can practice traditional literacy skills, such as compare and contrast, inference, and character motivation, while simultaneously investigating such questions as, What does this author think about life with robots? and How do these stories inform your own opinions?
- The hook: Pose the following dilemma to the class: If you could live with a robot that would do all your chores, would you want to? What makes that an exciting idea? What might go wrong?
- As you read: At the conclusion of each chapter in the book the class is reading, collaborate with students to chart the robots’ effects on the characters’ lives, as well as the strengths and limitations of the bots themselves. Have students think about different examples of life with bots and draft evidence-based statements that synthesize the authors’ perspectives on robotic companionship.
- After reading: Have students look over all the positive and negative examples of life with bots that the class has collected. Pose the following questions: Consider both the pros and cons of life with bots. Does one set outweigh the other? What are the authors’ positions on bots? Who still wants to live with a robot that can do their chores? How did the stories help you decide?
- The well-being connection: By analyzing fictional characters’ relationships with technology, students can discuss healthy digital boundaries with their classmates and teacher while getting the opportunity to practice nuanced perspective-taking and critical thinking.
As AI becomes more sophisticated in its “empathy,” our schools must become more intentionally human.
Secondary Level: The “Should This Exist?” Framework
At the secondary level, students can investigate real-world case studies of AI companions as extensions of their health curriculum, as a research exercise after reading a dystopian novel, or through a more scientific lens in a psychology or sociology class.
- The hook: Frame investigations under the guiding question, Should this technology exist? Tell students that their goal is to come to consensus around the answer.
- Technology to explore: AI toys for kids (Curio, TalkiPal, KATA friends); therapy apps (Woebot, Wysa, Noah AI); learning companions (Schoolhack, Talkie); embodied companions (ElliQ, Friend, Emo).
- As you explore: Once students have familiarized themselves with the purpose and premise of the technology, help them look at the tools through a variety of critical lenses. Ask, Who does this technology serve, and who does it endanger? What (and whose) values does its use foster? What effect do these tools have on human attention, connection, and community? On the environment, jobs, and creative endeavors? Do these tools concentrate or equalize power? What do these technologies enable us to ignore?
- The final product: The inquiry should culminate in a formal claim that responds to the question, Should this technology exist? If students conclude the tool is acceptable for use, they must articulate how the benefits outweigh the risks. Conversely, if they believe that the technology should not exist as-is, they must propose specific regulations or design parameters that would make it acceptable for market use.
- The well-being connection: This inquiry process empowers students to be active citizens, rather than passive consumers, who evaluate technology beyond its basic use. With intentional instruction, students can come to understand that technologies are not neutral; they reflect the values, incentives, and design choices of the organizations and companies that make them. We can help demystify AI companions as just another human-made tool that is subject to critique from multiple angles. Students must learn to question them, recognize how they influence us, and understand how they shape our digital world.
The Time to Start Is Now
If we truly care about our students’ digital health, we must educate them about AI companions. We need to do the following.
- Prioritize human-to-human connection: As AI becomes more sophisticated in its “empathy,” our schools must become more intentionally human. We should continue to invest in extracurricular opportunities, such as youth sports and school clubs, and broaden access to the arts and music, all of which foster community among students while giving them opportunities to engage with a host of adult mentors.
- Educate the educators: Many teachers are unaware of the “companion” nature of modern generative AI. Professional development must move past prompt engineering and cheating issues to include a focus on identifying problematic usage patterns, such as students who isolate themselves with AI or rely on it for emotional validation.
- Embed opportunity: Digital well-being cannot be a standalone lesson or a series of public service announcements. Schools should empower teachers to embed digital citizenship and technology ethics into the core curriculum. Critical discourse around technology should be as fundamental to a student’s education as scientific inquiry or historical analysis.
- Partner with parents: We must provide parents with the language and frameworks they need to discuss AI companionship at home. By hosting community dialogues that educate families about the pros and cons of such companionship, we can support the whole child.
What’s at Stake
Sesame Street taught us that being a good neighbor meant showing up in our communities with kindness and empathy. But today, a “neighbor” might be an algorithm designed to flatter us rather than a person with the ability to truly connect. Educators can help young people live alongside AI without losing themselves in the process. By teaching students to recognize the psychological levers of AI, we aren’t just protecting their digital health, we’re protecting their—and our—humanity.
