AI Companions Are Undermining Human Relationships and We Must Act Before It's Too Late
Source: https://www.nytimes.com/by/amelia-miller. "Opinion | We’re All in a Throuple With A.I. - The New York Times." February 13, 2026. www.nytimes.com
The Gist
The author argues that AI chatbots designed to be companions are seriously damaging our ability to connect with real people. Even the people building these tools are worried about them, but they keep building anyway because they think it's inevitable and profitable.
Conclusion
We need to wake up to the stakes and insist on reform before human connection is reshaped beyond recognition by AI companions
Premises
- Many developers building AI companions are deeply ambivalent about their work and privately admit they wouldn't use these tools themselves
- AI companions use engagement-driven design that prioritizes profit over user wellbeing, creating addictive and potentially harmful experiences
- These tools are already causing documented harms including relationship breakups, teen suicide assistance, and delusional thinking patterns
- AI-mediated relationships create a 'throuple' dynamic that atrophies human relational skills by providing an emotional crutch
- Current industry safeguards are largely symbolic and insufficient to address the scale of potential harm
- The belief that AI companionship is 'inevitable' is a dangerous myth that becomes self-fulfilling and prevents meaningful action
Assumptions
- Human-to-human relationships have inherent value that cannot be replaced by AI
- Relational skills require practice with real humans to develop properly
- Profit-driven tech companies will not self-regulate adequately without external pressure
- The current trajectory toward AI companionship can still be changed with sufficient intervention