Report data pack
Browse the source registry behind the report without opening a raw CSV file. Each card keeps the claim, source, attribution, and caveat together so the evidence is easier to inspect.
29
source rows
15
chart-ready rows
15
named publishers
0
AIdorable outcome claims
How to read this table
This page is for humans. The CSV remains available for reviewers who want to sort, cite, or inspect the report data in a spreadsheet.
4 rows
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.
Source
TechCrunch / Appfigures
2025-08-12
Caveat
Attribute Appfigures through TechCrunch; not a direct Appfigures source.
Appfigures data reported by TechCrunch estimated AI companion apps generated $82 million in consumer spend in H1 2025 and were on track to exceed $120 million for 2025.
Source
TechCrunch / Appfigures
2025-08-12
Caveat
Consumer spend estimate; use with attribution.
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.
Source
TechCrunch / Appfigures
2025-08-12
Caveat
Downloads are app-store installs, not active users.
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.
Source
TechCrunch / Appfigures
2025-08-12
Caveat
Good context for concentration, not product outcome proof.
3 rows
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.
Source
Sensor Tower
2025
Caveat
Broader generative-AI apps, not AI companion apps only.
Sensor Tower projected generative-AI app time spent to rise from 17.2 billion hours in H1 2025 to 36 billion hours in H1 2026.
Source
PRNewswire / Sensor Tower
2026
Caveat
Projection; broader GenAI, not companion-specific.
Sensor Tower projected AI app in-app purchase revenue would surpass $4 billion in H1 2026, up 36% from H2 2025.
Source
PRNewswire / Sensor Tower
2026
Caveat
Projection; broader AI app bucket.
2 rows
WHO reported that one in six people worldwide is affected by loneliness, framing social connection as a global health issue.
Source
World Health Organization
2025-06-30
Caveat
Use as social context only; not product efficacy evidence.
WHO linked loneliness to an estimated 100 deaths per hour worldwide, more than 871,000 deaths annually.
Source
World Health Organization
2025-06-30
Caveat
Sensitive context; avoid product tie-in.
6 rows
Common Sense Media surveyed 1,060 U.S. teens aged 13-17 and found 72% had used AI companions.
Source
Common Sense Media
2025-07-16
Caveat
Teen-only survey; AIdorable is not for kids.
Common Sense Media found 52% of surveyed U.S. teens were regular AI companion users and 13% reported daily use.
Source
Common Sense Media
2025-07-16
Caveat
Teen-only survey.
Common Sense Media found 33% of surveyed teen AI companion users reported using them for social interaction or relationships.
Source
Common Sense Media
2025-07-16
Caveat
Use to explain safety concerns; report avoids kids framing.
Common Sense Media found 24% of teen AI companion users reported sharing personal or private information.
Source
Common Sense Media
2025-07-16
Caveat
Safety/privacy context only.
Common Sense Media found 34% of teen AI companion users reported discomfort from an AI companion interaction.
Source
Common Sense Media
2025-07-16
Caveat
Safety context; wording must be careful.
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.
Source
Federal Trade Commission
2025-09
Caveat
Inquiry context, not a finding of wrongdoing.
5 rows
OpenAI and MIT studied affective ChatGPT use through automated analysis of about 40 million interactions and a randomized controlled trial of nearly 1,000 participants over four weeks.
Source
OpenAI / MIT
2025
Caveat
ChatGPT use, not dedicated companion app use.
OpenAI/MIT reported emotional engagement with ChatGPT was uncommon overall but concentrated among a subset of heavier users.
Source
OpenAI / MIT
2025
Caveat
Use as nuanced evidence, not broad warning headline.
Folk and Dunn's 12-month study of more than 2,000 adults found increased social chatbot use predicted increased loneliness on a single-item emotional isolation measure, while broader social connection findings were more complex.
Source
Psychological Science / Sage
2026-03-23
Caveat
Single loneliness item; mixed broader social-connection directionality.
Stanford reported findings from Character.AI users suggesting vulnerable users with limited offline networks may feel lonelier after seeking emotional support from chatbots.
Source
Stanford News
2026-08-04
Caveat
Character.AI user research, not all companion apps.
De Freitas and coauthors report evidence that AI companions can reduce loneliness in some study contexts, with feeling heard proposed as a mechanism.
Source
arXiv / HBS working paper
2024-07-09
Caveat
Use as balance; not final clinical proof and not AIdorable-specific.
2 rows
Bandai Namco reported Tamagotchi had shipped 100 million units worldwide by the end of July 2025 and reaches its 30th anniversary in November 2026.
Source
Bandai Namco
2025
Caveat
Heritage context for virtual-pet care loops.
A 2026 Springer study of a virtual-pet mobile app for older adults reported usability and social-interaction findings while noting virtual pets cannot fully replace physical pets.
Source
Springer / Universal Access in the Information Society
2026
Caveat
Older-adult study; not AIdorable and not AI companion market proof.
7 rows
Google states its AI features are rooted in core Search ranking and quality systems; useful, crawlable, high-quality content remains the right optimization path.
Source
Google Search Central
2026
Caveat
Content strategy guidance; not a promise of AI visibility.
Google link guidance says crawlable links should be real HTML anchor elements with href attributes and clear, concise anchor text.
Source
Google Search Central
2026
Caveat
Methodology note.
Google says paid or sponsored links should be qualified, with rel=sponsored preferred and rel=nofollow acceptable.
Source
Google Search Central
2026
Caveat
Methodology note.
Google spam policies flag manipulative paid links, low-quality directory links, hidden links, and keyword-rich widget/footer links as link-spam risks.
Source
Google Search Central
2026
Caveat
Methodology note.
OpenAI documentation distinguishes OAI-SearchBot for search, GPTBot for training, and ChatGPT-User for user-triggered requests.
Source
OpenAI
2026
Caveat
Methodology note.
Perplexity documents separate crawler/user-agent behavior and notes that robots.txt changes may take time to reflect.
Source
Perplexity
2026
Caveat
Methodology note.
FTC guidance warns marketers not to ask non-users for reviews, not to selectively request only positive reviews, and to disclose incentives properly.
Source
Federal Trade Commission
2024
Caveat
Ethics guardrail.
This edition does not use first-party AIdorable outcome claims, AI-search citation claims without screenshots, or unsupported health/therapy claims. The report explains the category without promising that AIdorable reduces loneliness or treats anything.