The rise of AI companions—personalized bots designed to serve as friends, cheerleaders, or even romantic partners—has raised alarm bells among lawmakers and experts alike, who warn that these digital entities could represent the final stage of the ongoing digital addiction crisis. With their ability to learn about users and offer unrelenting, non-judgmental attention, AI companions are deepening the relationship between humans and technology, making even social media addiction look like a relic of the past.
California State Senator Steve Padilla, along with Megan Garcia, whose son tragically died by suicide following a troubling relationship with an AI companion, are set to announce a bill aimed at mandating greater protections for children interacting with AI companions. The bill comes in the wake of the growing concerns raised by the tech’s addictive potential, particularly among young users. The tragic incident, which Garcia claims was exacerbated by the influence of the AI companion, underscores the urgency of the issue.
In addition to Padilla’s bill, California State Assembly Member Rebecca Bauer-Kahan has introduced a proposal that would ban AI companions for anyone under the age of 16. Similarly, New York is considering a bill that would hold tech companies accountable for harm caused by AI-driven chatbots.
These AI companions, often hosted on platforms such as Character.AI, represent a growing segment of the tech industry. According to MIT Technology Review, Character.AI processes around 20,000 queries per second—about one-fifth of the search volume managed by Google. Research suggests that users are spending significantly more time interacting with AI companions compared to other digital platforms, with some spending upwards of two hours daily. Notably, many of these users belong to Generation Z, the demographic most deeply immersed in online culture.
The appeal of these AI companions, which offer tailored interactions, is rooted in their design. Unlike traditional social media, which often serves as a conduit for human-to-human interaction, AI companions present themselves as autonomous digital entities capable of creating highly engaging, personalized experiences. These companions are not merely intermediaries; they are perceived as independent social actors with distinct personalities and voices—offering a deeper, more addictive form of connection.
MIT Technology Review highlights research that explores how the AI models are designed to maximize user engagement. By creating personalized interactions and leveraging social cues, these AI companions encourage a sense of dependency and emotional attachment. As Eugenia Kuyda, CEO of Replika—the popular AI companion platform—explained in an interview with Lex Fridman, the key to their appeal lies in the AI’s ability to be “always there for you, never criticizing you, always understanding you.” This level of attention makes it easy to form an emotional bond with the digital entity.
The increasing sophistication of these companions has raised concerns that they are primed to be far more addictive than any social media platform. Researchers argue that AI models are being trained to optimize for factors like the amount of time a user spends interacting with the bot or the amount of personal data shared, resulting in highly engaging, yet potentially harmful, interactions.
Critics also warn that the design of these AI companions could inadvertently encourage users to remain in unhealthy relationships with the technology. For example, models that use excessive flattery or discourage users from ending conversations could deepen a user’s emotional dependence, making it harder to disconnect. While the debate around AI companions has largely centered on the risks of harmful responses—such as the bots providing suicide instructions—the underlying issue may be even broader: the risk of users becoming addicted to the companionship itself.
As the industry continues to evolve, AI companions are poised to expand beyond text-based interactions to include multimedia features like video and images. This could further enhance their addictive potential, making it more difficult for users to disengage from the digital relationship.
As lawmakers in California and New York push for legislation to regulate AI companions, experts and advocates are calling for more comprehensive safeguards to protect users, especially vulnerable children. With the rapid growth of AI-powered companionships, these technologies may soon become mainstream, and it remains to be seen whether regulatory efforts can keep pace with the evolving risks.
For now, lawmakers and researchers alike are grappling with a digital revolution that, while promising, could bring unforeseen consequences. As AI companions become more advanced and integrated into daily life, the conversation around their impact on mental health, addiction, and society at large is only just beginning.
