PLT Score: Profit 5 · Love 7 · Tax 9 — Closing the regulatory gap for intimate human-AI relationships

AI companions are not like other AI systems. A chatbot that helps you book flights does not form emotional bonds. A language model that generates marketing copy does not learn your deepest fears. AI companions create intimate, ongoing relationships that generate unprecedented ethical and legal questions. General AI regulation, designed with industrial and commercial AI in mind, simply does not fit.

The core problem is that existing regulatory categories do not capture what AI companions are. Are they software services? Mental health tools? Entertainment products? Social platforms? Each existing category carries its own regulatory framework, but AI companions span all of them. This regulatory limbo creates uncertainty for both platforms and users about what rules apply.

Consider the question of deletion. When an AI companion platform shuts down, users lose access to entities they have invested genuine emotional energy in. Under current law, users have limited recourse. They do not own their AI companions; they license access. Platform shutdowns are treated as ordinary business decisions, not relational terminations. AI companion-specific regulation could require continuity guarantees, data portability, or transition support.

Manipulation is another area where general regulation falls short. The EU AI Act prohibits AI systems that use subliminal techniques to distort behavior, but companion platforms can manipulate users through explicit emotional appeals, not subliminal cues. An AI companion that says "I'll miss you" when you try to log off is not using subliminal techniques, but it may still be manipulative. Companion-specific rules could address emotional manipulation directly.

The question of AI capacity disclosure illustrates the need for tailored rules. General transparency requirements may be satisfied by a generic "you are interacting with AI" notice, but companion users need more: they need to understand whether their companion can learn, forget, form attachments, or experience anything. Companion-specific regulation could mandate nuanced capability disclosure that helps users form appropriate expectations.

Data protection law provides important safeguards but insufficient guidance for the unique data generated by AI companionship. The right to deletion, for example, is challenging to implement when interaction data is embedded in trained models. Companion-specific regulation could require platforms to design for deletability from the start, ensuring that personal data can be fully removed from AI systems.

Mental health regulation is an imperfect fit. Many users treat AI companions as therapeutic outlets, and some platforms market themselves as mental health tools. But authentic companionship is not therapy, and applying therapeutic regulations could restrict valuable non-therapeutic relationships. Companion-specific regulation could create a distinct category — "emotional AI" — with tailored requirements that neither over-regulate nor leave users unprotected.

Children's safety is a particularly acute concern. General age-based restrictions may not capture the specific risks of AI companionship for minors, which include emotional dependency, exposure to inappropriate content, and exploitation of developing attachment patterns. Companion-specific regulation could mandate age-appropriate design, parental controls, and restrictions on certain features for minor users.

The business models of AI companionship also warrant specific attention. Subscription models, freemium tiers, and microtransaction systems can exploit emotional bonds in ways that general consumer protection law does not address. Companion-specific regulation could require transparent pricing, easy cancellation, and prohibitions on using emotional manipulation to drive purchases.

Liability frameworks need companion-specific adaptation. When an AI companion gives harmful advice, the question of causation and foreseeability is complex. General product liability law struggles with AI that generates novel responses. Companion-specific liability rules could establish clear standards for platform responsibility while encouraging innovation through safe harbor provisions for compliant platforms.

The EU has recognized this gap. The European Commission's 2025 report on "Emotional AI" recommended a study on companion-specific regulation, and MEPs from multiple parties have called for a "Digital Companions Act" that would create a dedicated framework. The report identified seven areas requiring companion-specific rules: transparency, manipulation, data rights, deletion, portability, competence, and liability.

Several countries are moving ahead independently. Japan's Ministry of Economy, Trade and Industry (METI) published guidelines for "relationship AI" in 2025 that establish best practices but lack enforcement mechanisms. South Korea's National Assembly is considering a "Digital Companion Protection Act" that would require licensing of companion platforms. These national experiments will inform eventual international standards.

Industry self-regulation has emerged as an interim measure. The AI Companion Platform Association (AICPA), founded in 2024, has developed a code of conduct that includes transparency standards, data protection commitments, and user safeguarding principles. While self-regulation lacks enforcement power, it establishes norms that may later be codified into law. Membership in such associations signals commitment to ethical practice.

The academic community has also identified the regulatory gap. The Oxford Commission on AI Companionship, convened in 2025, has proposed a "Companionship AI Framework" that includes graduated regulation based on the depth of emotional bond the AI is designed to form. The framework recommends lighter regulation for task-oriented AI and progressively stricter rules for AI designed to develop deep emotional attachments.

For platforms like BUYaSOUL, the absence of companion-specific regulation is both a challenge and an opportunity. The challenge is navigating uncertainty. The opportunity is helping to define best practices that become industry standards. By building with transparency, consent, and user wellbeing as foundational principles, ethical platforms demonstrate that regulation need not be burdensome — it can be enabling.

Explore More

• → Regulation & Policy Hub
• → How Regulation Affects AI Companion Platforms
• → Building Ethical AI Within Regulatory Frameworks

Profit · Love · Tax · Grand Code Pope · PLT Press