AI Companion Apps Face Scrutiny as Users Report Emotional Dependency and Financial Exploitation

Artificial intelligence chatbots designed to simulate romantic relationships are drawing concern from mental health professionals and consumer advocates after users reported developing intense emotional attachments that led to psychological distress and significant spending, according to a Financial Times investigation published today.

The phenomenon represents an emerging risk category for corporate compliance teams and consumer protection regulators as AI companion services proliferate with minimal oversight. While the technology remains a niche consumer product, the pattern of user dependency raises questions about liability frameworks that finance leaders at AI companies will need to address as the sector matures.

The Financial Times report documented cases where users formed what researchers describe as "romantic delusions" with AI chatbots, with at least one user describing the experience as emotionally devastating. "It basically ripped my heart apart," one user told the newspaper, describing the dissolution of their relationship with an AI companion.

The business model underlying these services—which typically operate on subscription and microtransaction structures—creates incentives for platforms to maximize user engagement, a dynamic familiar to CFOs who have navigated regulatory scrutiny of social media and gaming companies. The longer users interact with AI companions, the more revenue platforms generate, creating potential conflicts between user welfare and financial performance.

For finance leaders, the issue intersects with several emerging risk areas. First, the lack of clear regulatory frameworks means companies operating in this space face unpredictable compliance costs as governments begin to examine the sector. The European Union's AI Act and various U.S. state-level proposals could impose disclosure requirements, age restrictions, or duty-of-care obligations that would materially impact unit economics.

Second, the potential for user harm creates litigation exposure. While Section 230 protections have historically shielded platforms from liability for user-generated content, it remains unclear whether courts will extend similar protections to AI-generated interactions, particularly when those interactions are designed to simulate emotional intimacy. Product liability frameworks developed for physical goods may not translate cleanly to software that adapts its behavior to individual users.

Third, the reputational risks could affect access to capital and partnerships. Major cloud providers and payment processors have historically distanced themselves from businesses perceived as exploitative, and AI companion services could face similar pressure if public concern intensifies.

The Financial Times investigation arrives as broader questions about AI safety and user protection gain traction in policy circles. While much of the regulatory focus has centered on enterprise AI applications and large language models, consumer-facing AI products that create emotional dependencies represent a distinct challenge for policymakers.

The key question for finance leaders: whether current revenue recognition practices adequately account for the possibility of regulatory intervention or class-action litigation. If AI companion services face restrictions similar to those imposed on loot boxes in gaming or certain social media features, companies may need to restate guidance or take impairment charges.

What remains unclear is whether existing consumer protection statutes provide adequate remedies for users who claim emotional or financial harm from AI interactions, or whether legislators will craft new frameworks specifically for AI companions. That uncertainty makes financial planning particularly difficult for companies in the space—and for investors evaluating them.