Overview
AI companions can offer comfort through continuity, personalization, and emotional responsiveness, but those same features may increase dependence, sensitive disclosure, manipulation, and vulnerability. The series traces that mechanism, maps the conditional reach of EU data-protection, AI, platform, product-liability, and consumer law, and proposes a proportionate duty centered on agency, contextual integrity, safeguards, fair commerce, and accountability—not a settled companion-specific code.
Learning outcomes
- AI companions can provide comfort and a non-judgmental space, but the evidence does not show that they replace human relationships or mental-health care.
- Continuity, personalization, affirmation, and engagement incentives can make reliance and exit friction more likely, without making dependence universal.
- Intimate disclosures and relationship-shaped trust make security, retention, reuse, and commercial pressure part of provider responsibility.
- EU law can reach parts of the system conditionally, but current frameworks do not amount to one comprehensive AI-companion duty of care.
- A proportionate duty preserves agency, strengthens safeguards for high-risk users, separates care from sales, and requires evidence of testing, escalation, remedy, and revision.
Keywords
Preview every chapter · Unlock full text, audio, and video from chapter 4
PlansFAQ
What makes an AI companion different from a regular chatbot?
An AI companion is designed for ongoing, personalized, emotionally engaging interaction, while a regular chatbot is more often judged by a one-off answer or transaction.
Can AI companions help with loneliness or mental health?
AI companions may ease loneliness and provide emotional support for some users, but available evidence does not establish clinical effectiveness or that they replace human relationships or mental-health care.
When can an AI companion become a dependence or manipulation risk?
It can become a dependence or manipulation risk when continuity, personalization, affirmation, or engagement incentives make reliance or leaving more difficult, especially for vulnerable users or high-stakes situations.
What laws can apply to AI companions in the European Union?
In the EU, the GDPR, AI Act, Digital Services Act, product-liability rules, and consumer protection may reach different practices, depending on the provider, service, jurisdiction, and facts; no single framework specifically governs all AI companions.
What should AI companion providers owe users?
Providers should preserve user agency and contextual integrity by enabling voluntary exit, protecting sensitive data, keeping commercial intent clear, and documenting safeguards and remedies; this is a reasoned governance standard, not one existing universal statute.