Market signal
337
The category is real
revenue-generating AI companion apps
A young market with measurable downloads, spend, and revenue concentration.
Appfigures via TechCrunch
Back to AIdorableView sources 
AIdorable Research · 2026 Edition
AI Companion AppStatistics 2026
The Digital Nurturing Report
AI companion apps are becoming a measurable consumer category. The useful report is not a hype piece: it separates companion-app growth from broader GenAI, labels every caveat, and keeps safety research in the foreground.
The frame AIdorable is building toward is digital nurturing: non-romantic, virtual-pet-style care loops built around daily rituals, memories, outfits, and milestones.
Visual thesis
Growth is the hook. Trust is the moat. Care is the category.
The report works when readers can understand the entire argument before they read a paragraph: companion apps are growing, the safety evidence is serious, and AIdorable belongs in the non-romantic care-loop lane.
01 · Market
Growth
02 · Safety
Boundaries
03 · Research
Caveats
04 · Nurturing
Positioning
Safety signal
72%
The boundaries matter
surveyed teens had used AI companions
Teen use, privacy sharing, discomfort, and regulatory attention make careful framing non-negotiable.
Common Sense Media
Product signal
100M
The care loop is different
Tamagotchi units shipped worldwide
Digital nurturing sits closer to virtual-pet ritual than romantic chatbot or therapy framing.
Bandai Namco
AIdorable care loop
Adopt → Care → Remember → Return
1
Adopt
Meet a tiny companion
2
Care
Tend through daily rituals
3
Remember
Build moments and continuity
4loop
Return
Come back as the bond grows
The product story is continuity: a user returns because the baby, memories, and milestones keep becoming more personal.
337
Revenue-generating AI companion apps
Appfigures via TechCrunch
$82M
AI companion app consumer spend in H1 2025
Appfigures via TechCrunch
220M
Estimated global downloads by July 2025
Appfigures via TechCrunch
72%
Surveyed U.S. teens who had used AI companions
Common Sense Media
100M
Tamagotchi units shipped worldwide
Bandai Namco, 2025
29
citation-ready claims in this edition
AIdorable source registry
Source base
Built from named public sources, not anonymous roundup stats.
WHO
social connection context
Common Sense
teen safety survey
TechCrunch / Appfigures
companion app market data
Sensor Tower
broader GenAI app context
Stanford
companion risk research
OpenAI / MIT
affective-use study
Growth
Market
Companion-specific downloads, spend, app supply, and broader GenAI context are separated so the numbers stay honest.
Jump to section →
Boundaries
Safety
Teen use, privacy concerns, and regulatory attention are framed as category risks, not as fear marketing.
Jump to section →
Mixed evidence
Research
Studies are presented by population, context, what they suggest, and what they do not prove.
Jump to section →
Care loop
Digital Nurturing
AIdorable belongs closer to virtual-pet-style daily care than romantic chatbot or therapy framing.
Jump to section →
Digital nurturing describes apps where the core loop is care. The product is not mainly about answering questions, flirting, or replacing people. It is about checking in, tending, naming, remembering, dressing, and watching something grow.
At a glance
The useful story is growth plus boundaries.
This page should make AIdorable look credible by being careful: the biggest numbers are attributed, every chart has caveats, and the product is kept out of therapy/kids/medical framing.
- Appfigures data reported by TechCrunch identified 337 active revenue-generating AI companion apps worldwide by mid-2025, with 128 released in 2025 by the report date.
- The same Appfigures data estimated AI companion apps generated $82 million in consumer spend in H1 2025 and were on track to exceed $120 million for 2025.
- Appfigures data reported by TechCrunch estimated AI companion apps had reached 220 million global downloads by July 2025, with H1 2025 downloads up 88% year over year to 60 million.
- Broader generative-AI apps neared 1.7 billion downloads and almost $1.9 billion in in-app purchase revenue in H1 2025, according to Sensor Tower — a separate, wider bucket than companion apps.
- WHO reported that one in six people worldwide is affected by loneliness. This is social context, not evidence that any one app improves health outcomes.
- Common Sense Media's teen research shows why AI companion products need sharp safety and privacy boundaries, especially around minors.
- Research evidence is mixed: some studies raise concerns about emotional dependence and loneliness for vulnerable users, while other work suggests AI companions may help some people feel heard in specific contexts.
- Virtual-pet care loops have deep consumer history: Bandai Namco reported Tamagotchi had shipped 100 million units worldwide by the end of July 2025.
Section 1
AI Companion Apps Are Growing Quickly
AI companion apps are now a measurable consumer software category. Appfigures data reported by TechCrunch identified 337 active revenue-generating AI companion apps worldwide by mid-2025. Of those, 128 had been released in 2025 by the report date.
The same Appfigures data estimated that AI companion apps generated $82 million in consumer spend in H1 2025 and were on track to exceed $120 million for the year. Downloads were also large: the category had reached an estimated 220 million global downloads by July 2025, and H1 2025 downloads were up 88% year over year to 60 million.
The category is also concentrated. Appfigures data reported by TechCrunch said the top 10% of AI companion apps generated 89% of category revenue, and about 33 apps had passed $1 million in lifetime consumer spend.
AI companion app demand, 2025
Downloads are app-store installs, not active users. Run-rate is a projection, not observed full-year revenue.
Global downloads by July 2025220M
H1 2025 downloads (up 88% YoY)60M
2025 consumer spend run-rate (projected)$120M+
H1 2025 consumer spend$82M
Supply and revenue concentration
The top 10% of AI companion apps generated 89% of category revenue by mid-2025.
Active revenue-generating AI companion apps (mid-2025)337
Released in 2025 by the report date128
Apps above $1M lifetime consumer spend~33
Broader generative-AI app context — not companion-specific data
The broader generative-AI app market gives additional context, but it should not be blended with companion-specific data. Sensor Tower reported that generative-AI apps neared 1.7 billion downloads and nearly $1.9 billion in in-app purchase revenue in H1 2025. Sensor Tower also projected generative-AI app time spent to rise from 17.2 billion hours in H1 2025 to 36 billion hours in H1 2026, with AI app in-app purchase revenue projected above $4 billion in H1 2026.
1.7B
GenAI app downloads, H1 2025
$1.9B
GenAI in-app purchase revenue, H1 2025
17.2B → 36B hrs
Time spent, H1 2025 → H1 2026 (projected)
$4B+
Projected AI app IAP revenue, H1 2026
Sources: Sensor Tower, State of AI Apps (H1 2025) · Sensor Tower State of AI 2026 report (via PRNewswire). Broader generative-AI apps, not AI companion apps only. Projections are not observed full-year facts.
Important distinction: AI companion app figures and broader generative-AI app figures are related, but they are not the same category. Companion-specific charts on this page use Appfigures-via-TechCrunch data; broader AI app context is labeled separately.
Section 3
Safety, Teens, And Regulation
AI companion products need a safety-aware frame because use among younger audiences is already visible. Common Sense Media surveyed 1,060 U.S. teens aged 13 to 17 and found 72% had used AI companions. The report also found 52% of surveyed teens were regular AI companion users, and 13% reported daily use.
The same survey found 33% of teen AI companion users reported using them for social interaction or relationships, 24% reported sharing personal or private information, and 34% reported discomfort from an AI companion interaction.
These are teen-only findings. They should not be generalized to adult users, and they should not be used to make claims about AIdorable. They do show why the category needs clear product boundaries, privacy language, and age framing.
Teen AI companion use — safety context
Survey of 1,060 U.S. teens aged 13–17. Teen-only findings; not generalizable to adults. AIdorable is not a kids app.
Teens who had used AI companions72%
Teens who were regular AI companion users52%
Teen users reporting discomfort from an interaction34%
Teen users using companions for social interaction or relationships33%
Teen users reporting sharing personal or private information24%
Teens reporting daily AI companion use13%
Source: Common Sense Media, Talk, Trust, and Trade-Offs (Jul 2025)
Regulatory signal
Regulators are paying attention too. The FTC opened an inquiry into AI companion chatbots and requested information from companies about safety, monetization, age restrictions, disclosures, negative impacts, and data handling. An inquiry is not a finding of wrongdoing, but it is a clear signal that AI companion products are being examined as a safety-sensitive category.
Source: Federal Trade Commission (Sep 2025) · Inquiry context, not a finding of wrongdoing.
Section 4
What Research Says About AI Companions
The evidence on AI companions is mixed and context-dependent. A serious report should not claim that AI companions are simply good or bad.
OpenAI / MIT affective use study
OpenAI / MIT Media Lab affective use study (2025)
Population / context
Automated analysis of ~40 million ChatGPT interactions plus a randomized controlled trial of nearly 1,000 participants over four weeks. General ChatGPT use, not dedicated companion apps.
What it suggests
Emotional engagement with ChatGPT was uncommon overall, but concentrated among a subset of heavier users.
What it does not prove
That AI companion apps broadly cause emotional dependence, or that most users engage emotionally with AI.
Folk & Dunn, Psychological Science
Folk & Dunn, Psychological Science (2026)
Population / context
12-month study of more than 2,000 adults measuring social chatbot use and social connection over time.
What it suggests
Increased social chatbot use predicted increased loneliness on a single-item emotional isolation measure; broader social connection findings were more complex.
What it does not prove
That all companion products increase loneliness for all users, or anything about non-chatbot care loops.
Stanford research on Character.AI users
Population / context
Findings from Character.AI users, a chatbot-style companion product with a large user base.
What it suggests
Vulnerable users with limited offline networks may feel lonelier after seeking emotional support from chatbots.
What it does not prove
That every companion app produces the same outcome. It is a product-specific finding, not an all-app finding.
De Freitas et al. (working paper)
De Freitas et al., working paper (arXiv)
Population / context
Working-paper studies of AI companion interactions, proposing "feeling heard" as a mechanism.
What it suggests
AI companions can reduce loneliness in some study contexts.
What it does not prove
Final clinical proof, generalized benefit claims, or anything specific to AIdorable.
Bottom line: design context matters. A chatbot used as an emotional substitute, a romantic companion app, a general assistant, and a virtual-pet-style care loop may all produce different user experiences. This report talks about those differences instead of flattening the whole category into one claim.
Section 5
Digital Nurturing And The Virtual-Pet Lineage
AI companion apps are often discussed as chatbots, but chat is not the only relevant history. Another lineage is the virtual pet: a digital being that invites care, routine, naming, dressing, returning, remembering, and watching it grow.
Category language
Digital nurturing is not just chatbotting.
The distinction matters because the report is not claiming every AI companion works the same way. A care loop asks users to return, tend, remember, and watch something grow.
Conversation loop
Typical chatbot framing
- Prompt and response
- Advice or emotional support framing
- Session-by-session conversation
Care loop
Digital nurturing framing
- Adopt and tend to a tiny companion
- Daily rituals, outfits, memories, and milestones
- Non-romantic continuity over time
100M
Tamagotchi units shipped worldwide by end of July 2025
Bandai Namco reported that Tamagotchi had shipped 100 million units worldwide by the end of July 2025 and reaches its 30th anniversary in November 2026. That does not prove anything about AIdorable outcomes, but it does show how durable the virtual-pet care loop has been in consumer culture.
Academic signal
Academic research also supports taking virtual-pet design seriously as a social interaction pattern. A 2026 Springer study of a virtual-pet mobile app for older adults reported usability and social-interaction findings while noting that virtual pets cannot fully replace physical pets. This is not AI companion market proof, and it is not an AIdorable study, but it helps explain why digital care loops deserve their own category language.
Source: Universal Access in the Information Society, Springer (2026)
Digital nurturingis the term this report uses for that category. It describes products whose primary emotional loop is care rather than flirtation, productivity, or advice. In AIdorable’s case, the frame is a non-romantic AI baby companion for cozy daily care: a small companion users can check in on, dress, remember with, and watch grow.
Section 6
What AIdorable Is And Is Not
AIdorable boundary box
A nurturing companion, not a therapy claim.
AIdorable is a non-romantic AI baby companion for cozy daily care. It uses a virtual-pet-style nurturing loop: users care for a tiny companion through daily rituals, memories, outfits, and milestones.
This report does not claim AIdorable reduces loneliness, improves mental health, provides therapy, or supports childcare or parenting decisions.
What we are
- AI companion
- Virtual pet
- Cozy daily care
- Digital nurturing
- Mobile lifestyle app
What we are not
- Therapy
- Medical support
- Childcare advice
- Parenting advice
- Kids app
- Romantic AI
- Crisis support
Section 7
Methodology
This report classifies sources before using them, then keeps the public data pack attached to the page so readers can inspect the evidence behind the presentation.
Explore the report data
A tidy data pack for journalists, builders, and reviewers.
Includes source URLs, claim notes, caveats, and chart values so readers can inspect the evidence behind the report instead of trusting a stat screenshot.
Citation-ready onlyUpdated 202610+ source rows
Explore sources
Browse source URLs, quote-safe claims, dates, notes, and caveats in a designed source page.
Explore chart data
Review the public chart values, source links, and caveats before downloading the spreadsheet.
The public CSVs include only citation-ready claims and chart rows used by this report edition.
Source trust method
A research report should show its receipts.
The methodology is built to keep the page useful without turning into an unsupported marketing claim. Source rows are sorted before they become page copy, chart data, or citation snippets.
FindClassifyCaveatPublish
Claim filter
Every stat is classified first.
Rows are marked citation-ready, lead-only, or reject before they can appear in public copy or charts.
Source context
Dates and caveats travel with the claim.
The source table keeps URLs, publication timing, claim notes, and limits so readers can inspect the context.
Chart hygiene
Charts use only citation-ready rows.
The public chart-data CSV excludes lead-only material, first-party outcome claims, and unsupported AI-citation claims.
Safety boundary
No AIdorable outcome claims.
The report does not claim AIdorable reduces loneliness, treats anything, or is for children.
citation-ready
A claim can be used in public copy with attribution and caveats.
lead-only
A claim may guide future research but is not quoted or charted in this edition.
reject
A claim is not usable in this report edition.
Included source types
- Official public-health and government sources
- Survey reports with clear population labels
- Peer-reviewed or university research summaries
- Industry data reported by named publishers
- Official search/crawler documentation for methodology notes
Excluded from this edition
- First-party AIdorable outcome claims
- AI-search citation claims without screenshots/query logs
- Quantitative APA or HBR claims without exact extraction
- Lead-only adjacent examples such as NYSOFA and Finch
- Generic stat roundups and affiliate listicles
Appendix
Search And Crawlability Notes
Google says its AI features are rooted in core Search ranking and quality systems, which means the durable optimization path is useful, crawlable, source-rich content. (Google Search Central: AI optimization guide)
Google’s link guidance says crawlable links should be real HTML anchor elements with href attributes and clear, concise anchor text. Paid or sponsored links should be qualified, with rel="sponsored" preferred and rel="nofollow" acceptable. Google’s spam policies flag manipulative paid links, low-quality directory links, hidden links, and keyword-rich widget/footer links as link-spam risks. (links best practices) · (qualify outbound links) · (spam policies)
For AI crawler posture, OpenAI documentation distinguishes OAI-SearchBot, GPTBot, and ChatGPT-User, while Perplexity documents separate crawler and user-agent behavior. Crawler access and good structure do not guarantee inclusion in AI answers. (OpenAI bots) · (Perplexity crawlers)
For journalists and researchers
Citation Snippets
These summaries are safe to quote with attribution to this report and the underlying sources linked above.
Short category summary
AI companion apps are growing quickly, but the research is mixed and safety-sensitive. AIdorable frames its product as digital nurturing: a non-romantic, virtual-pet-style care loop built around daily rituals, memories, outfits, and milestones.
Market summary
Appfigures data reported by TechCrunch estimated AI companion apps had reached 220 million global downloads by July 2025 and generated $82 million in consumer spend in H1 2025. Broader Sensor Tower data shows generative-AI apps nearing 1.7 billion downloads and almost $1.9 billion in in-app purchase revenue in H1 2025.
Safety summary
Common Sense Media's 2025 teen survey found widespread AI companion exposure among U.S. teens, including regular use, private-information sharing, and discomfort from some interactions. The report treats those findings as a reason for sharper age, privacy, and claim boundaries.
Digital nurturing summary
Digital nurturing describes apps where the core emotional loop is care: checking in, tending, naming, remembering, dressing, and watching something grow. It is closer to virtual-pet-style ritual than romantic chatbot framing.
Related Reading
What is a virtual baby app? →Virtual pet apps for adults, reviewed →Apps like Tamagotchi →Best virtual baby apps →Replika app review →AIdorable press kit and media assets →
Curious what digital nurturing feels like?
AIdorable is a non-romantic AI baby companion for cozy daily care — daily rituals, memories, outfits, and milestones.
Published by AIdorable · Last updated August 15, 2026. All third-party figures are attributed to their original publishers; caveats and population labels are part of each claim. This report makes no loneliness, mental-health, or therapeutic claims about AIdorable.
Section 2
Why Social Connection Is Part Of The Conversation
The rise of AI companion apps is happening against a wider social backdrop. WHO reported in 2025 that one in six people worldwide is affected by loneliness and linked loneliness to an estimated 100 deaths per hour worldwide, more than 871,000 deaths each year.
Those numbers matter because they explain why companion products attract attention. They do not prove that AI companion apps solve loneliness, and they do not prove anything about AIdorable outcomes.
The safer conclusion is narrower: social connection is a major public issue, and that makes emotional technology categories more sensitive. Products in this space should be precise about their claims, careful about user boundaries, and honest about what the evidence does and does not show.
1 in 6
people worldwide affected by loneliness, per WHO (2025)
871,000+
deaths per year linked by WHO to loneliness — about 100 deaths every hour
Source: World Health Organization (Jun 2025) · Context, not efficacy: these figures describe a public-health backdrop. They are not evidence that any app improves health outcomes.