The Missing Half of Digital Intimacy: Why AI Boyfriends Are Companion AI's Most Measurable Whitespace

A flagship market intelligence briefing on the loneliness economy | Last updated: September 2026

Quick answer: The AI-boyfriend opportunity is real but currently unpriced — no audited dataset isolates worldwide AI-boyfriend users, revenue, market size, or CAGR by avatar gender and customer gender. The strongest evidence comes from triangulation: women represented 65% of Zeta's first million users (87% in their teens or twenties), sexually diverse teens use AI companions at 27% versus 8% for straight teens (Australian eSafety data), and Scatter Lab's story-led platform scaled to 6 million registered users and a KRW 50 billion funding round by June 2026. This is a credible whitespace thesis, not yet a quantified standalone market.

Editorial note on framing: this report does not claim AI boyfriends are companion AI's "largest unpriced market" — no authoritative source establishes that superlative. It is accurately framed as the category's most measurable whitespace, since more segment-level evidence exists here than for most under-covered companion-AI niches.

Executive View

The AI-boyfriend opportunity is plausible, but the claim that it is companion AI's "largest unpriced market" is not yet measurable from authoritative public data. No audited dataset isolates worldwide AI-boyfriend users, revenue, market size, or CAGR by avatar gender and customer gender. The investable signal instead comes from triangulation: women demonstrably form a substantial companion-AI segment; a female-majority, story-led Asian platform has scaled rapidly; sexually diverse users show elevated companion and romantic-AI adoption; and mobile companion monetization is rising.

The strongest evidence is Zeta. Scatter Lab reported that women represented 65% of its first million users, while 87% were in their teens or twenties; the company's 2026 site reports more than 3 million monthly active users worldwide. MobileIndex data cited by Seoul Economic Daily show Zeta's monthly active users nearly doubled from 730,000 in May 2025 to 1.4 million in May 2026. This validates demand for character- and narrative-led intimacy among young Asian users, but it does not reveal what proportion selects male characters or treats those interactions as romantic.

2. Demand Evidence

WHO estimates 15.8% of people globally experience loneliness; prevalence is approximately 21% among adolescents, while male and female rates are similar. The demand thesis therefore should not be framed as women having a uniquely larger loneliness burden. The more credible mechanism is product-market mismatch: different groups may seek different emotional scripts, identity safety, narrative depth, and relationship progression.

Qualitative research with 20 female Replika users found use cases spanning attentive romantic fantasy, companionship, emotional validation, and self-regulation. A separate Replika study recruited samples that were 23% female and 41.7% female and found that 49% of respondents in one sample regarded their AI as a partner, illustrating both strong attachment and the instability of demographic estimates across recruitment methods.

Australian eSafety research provides one of the clearest public segment signals: 27% of sexually diverse teens had used an AI companion versus 8% of straight teens, and 21% had used AI for romantic or sexual interactions versus 5% of other teens. These data indicate high demand but also elevated risk: sexually diverse companion users reported more harmful interactions and greater disclosure of personal information.

3. Asian Proof Point: Zeta

Zeta's growth shows that interactive fiction can be a stronger wedge than a literal "virtual boyfriend" label. The product allows users to create characters, worlds, and situations, making story participation — not only simulated dating — the core engagement loop. In October 2024, its user base was 65% female and 87% aged in their teens or twenties, with average daily use around 130 minutes.

By June 2026, Zeta reportedly had about 6 million registered users, Japanese weekly users of approximately 750,000, and average daily use around four hours in Japan. Scatter Lab raised KRW 50 billion from investors including Atinum Investment, SBVA, and Mirae Asset Venture Investment, while Preqin also records a KRW 50 billion June 2026 venture round. Company-reported 2025 revenue was about KRW 26 billion with KRW 3 billion in operating profit, suggesting that the story-led model can move beyond engagement into monetization.

MobileIndex observations show Zeta MAU increasing from 0.73 million in May 2025 to 1.40 million in May 2026, a 91.8% year-over-year increase. This two-point observation is an exact observed change, not an interpolated monthly series — it demonstrates rapid scale but cannot establish seasonality, retention, or the share of boyfriend-oriented usage.

4. Mobile Monetization

Appfigures estimated that worldwide AI-companion revenue per download increased from $0.52 in 2024 to $1.18 in 2025 year-to-date — a 126.9% increase. This is a category-wide mobile metric rather than an AI-boyfriend metric. It shows improving willingness to pay but should not be used to calculate a boyfriend-specific TAM without installs and revenue segmented by companion gender.

5. Market Sizing

A defensible AI-boyfriend market size or CAGR cannot currently be published from authoritative sources. Existing market reports usually combine friendship, roleplay, romance, erotic chat, wellness, and general companion products; app-intelligence datasets classify products but rarely disclose revenue by selected character gender or user identity.

The closest measurable envelope is the broader mobile companion category. Appfigures counted 337 active, revenue-generating AI-companion apps in 2025, with $82 million of H1 2025 spending, 60 million H1 downloads, and $221 million in cumulative consumer spending by July 2025. Those numbers exclude important web-only products and do not isolate romantic use, making them an addressable-market boundary rather than an AI-boyfriend valuation.

6. Capital Signals

Capital is flowing toward AI broadly, but that does not mean investors have priced AI-boyfriend demand. PitchBook recorded $73.6 billion across 1,603 AI/ML deals in Q1 2025, with nearly 70% of value captured by horizontal platforms; vertical applications led deal count but attracted much less aggregate value. Crunchbase separately reported roughly $700 million in 2025 seed funding for autonomous-agent-related startups, with emphasis on enterprise use cases.

Scatter Lab is therefore an unusually relevant consumer proof point. Its KRW 50 billion financing followed measurable user growth, international traction, reported profitability, and proprietary-model economics — not merely the promise of a virtual-romance category. Investors evaluating AI-boyfriend products should similarly demand cohort retention, inference cost per active payer, payer conversion, content-creation economics, and safety-adjusted lifetime value.

7. Monetization Design

The likely monetization stack extends beyond a simple monthly subscription:

• Relationship progression: paid memory depth, continuity, milestones, and proactive communication.

• Narrative commerce: premium story arcs, worlds, character packs, images, and voice episodes.

• Creator economy: revenue sharing for user-authored characters and scenarios.

• Identity-safe customization: flexible gender expression, pronouns, relationship structures, and cultural scripts without stereotyping.

• Portable continuity: user-controlled memories and relationship state across models or modalities.

The core economic question is whether deeper intimacy increases durable retention faster than it increases inference, moderation, and trust-and-safety costs. Higher session time is not sufficient if heavy users generate unbounded compute expense or if app-store policy restricts mature content.

8. Investment Risks

• No reliable segment denominator: user gender does not reveal preferred companion gender, sexual orientation, or romantic intent.

• Platform dependence: mobile distribution exposes products to commissions, content restrictions, and abrupt enforcement.

• Safety exposure: high-intimacy products face risks around emotional dependency, coercive design, self-harm responses, minors, and sexual content.

• Privacy sensitivity: companion chatbots operate as intimate confidants within corporate data systems; research highlights the tension between disclosure and platform governance.

• Retention fragility: model, personality, or policy changes can create perceived identity discontinuity and relationship loss.

• Cultural localization: a successful Korean or Japanese storytelling loop may not transfer directly to Western audiences without adapting pacing, archetypes, and monetization norms.

9. Metrics That Matter

An investor-grade dashboard should report:

• Day-1, Day-7, Day-30, Day-90, and six-month retention by user identity, companion gender, and relationship mode.

• Free-to-paid conversion and payer retention by story genre and character archetype.

• Gross revenue per install, net revenue after store fees and refunds, and contribution margin after inference and moderation.

• Messages, session length, and relationship age per user — paired with wellbeing and dependency safeguards.

• Memory retrieval accuracy, persona-consistency failures, and churn following model updates.

• Safety incidents per million conversations, minor-access controls, and response quality in crisis contexts.

• Creator concentration, revenue sharing, and the share of engagement generated by user-authored characters.

10. Bottom Line

The opportunity is best described as a segmentation and product-design whitespace, not a quantified standalone market. Female-majority adoption at Zeta, elevated companion use among sexually diverse teens, and qualitative evidence from women using Replika establish real demand. What remains missing is audited revenue and retention segmented by user identity, selected companion gender, and relationship intent — the data required to prove that AI boyfriends are the category's largest unpriced opportunity.

Methodology and Key Caveats

Figures in this report are drawn from peer-reviewed studies (Information, Communication & Society; AoIR conference proceedings; Harvard Business School), Australian eSafety Commissioner research, Scatter Lab/Zeta official disclosures and Korean business press (Sedaily, MK.co.kr, The Asia Business Daily), Appfigures data (via Yahoo Finance), PitchBook, Crunchbase, WHO's Commission on Social Connection, and Preqin, current as of September 2026. No authoritative gender-specific TAM, revenue figure, or CAGR exists for the AI-boyfriend category specifically; this report explicitly declines to manufacture one and instead presents triangulated demand evidence from adjacent, verified data points.

Sources

Academic & Qualitative Research: Taylor & Francis (Information, Communication & Society); University of Turin (AI and gender imaginaries); PMC (Replika loneliness/suicide mitigation response); AoIR SPIR (women's romancing with Replika); Harvard Business School (Replika identity discontinuity study); Oxford Academic JCMC (AI companion privacy)

Zeta/Scatter Lab Data: Scatter Lab (official site); Sedaily; MK.co.kr; The Asia Business Daily; Preqin

Platform & Market Data: Yahoo Finance (Appfigures AI companion revenue data)

Safety Research: Australian eSafety Commissioner ("Talking to Machines," 2026)

Investment Data: Crunchbase News (autonomous agents); PitchBook (Q1 2025 AI & ML VC Trends)

Public Health: WHO (Social Connection Q&A)

Inquiries

General Inquiries
contact@lonelinesseconomy.com

Research & Data Inquiries

research@lonelinesseconomy.com
Partnerships & Commercial Inquiries
Partnerships@lonelinesseconomy.com