165 Million People Visit This AI Chat App Monthly — It's Also Facing Wrongful Death Lawsuits
Character.AI Review 2026: pricing, product, and safety record — including the first wrongful death settlements against an AI company in U.S. history, settled in January 2026.
📋 Table of Contents
Character.AI draws roughly 163 million visits every three months — more web traffic than Perplexity or Claude — with users averaging 14 minutes per session and nearly 10 pages browsed each visit, engagement numbers most social platforms would envy. Launched in September 2022 by former Google engineers Noam Shazeer and Daniel de Freitas, the platform let users chat with AI-driven characters built from fiction, history, or their own imagination, and it grew fast: 28 million monthly active users and a $2.5 billion valuation by mid-2024.
That's one half of the story. The other half is that Character.AI has lost roughly 8 million of those monthly users, around 60% of its peak valuation, both of its founders, and — in January 2026 — became the first AI company in U.S. history to settle wrongful death lawsuits, following the suicides of two teenagers who had extensive, emotionally intense interactions with the platform's chatbots. Both halves are true at the same time, and a fair review has to hold them together rather than picking one.
What Character.AI Is
The core product is simple to describe: pick or create a character — a fictional figure, a historical person, a celebrity persona, or something entirely original — and have an open-ended conversation with it. The character library exceeds 10 million user-created entries, and the platform runs on web, iOS, and Android across 31 languages, with the mobile apps considerably more feature-complete than the browser version.
The audience skews young: the largest age group is 18-24, and the platform's historical user base has run heavily 16-30. Direct traffic — people typing the URL or opening the app rather than arriving through search — sits above 70%, a sign of strong habitual return behavior rather than casual discovery.
The Google Deal
In August 2024, Google paid roughly $2.7 billion to license Character.AI's underlying model technology and bring on its two founders, along with about 30 key researchers, to work on Gemini. The deal was structured as a non-exclusive technology license rather than an acquisition — it didn't change who owned Character.AI — but its practical effect on the company was severe. It lost the people who built its core technology, and it stopped developing its own proprietary models, shifting instead to open-source backbones from DeepSeek and Meta.
That $2.7 billion was a one-time payment, not recurring revenue — it extended the company's runway without solving its underlying business model. In a notable postscript, Noam Shazeer announced in June 2026 that he was leaving Google for OpenAI, closing a loop that started with him leaving Google to found Character.AI in the first place.
Pricing & Monetization
| Plan | Price | Includes |
|---|---|---|
| Free | $0 | Full character library, basic model, mid-chat ads, daily usage limits, limited voice calls |
| c.ai+ Monthly | $9.99/month | Ad-free, better model, unlimited regenerations and voice calls, enhanced memory |
| c.ai+ Annual | $94.99/year (~$7.92/mo) | Same as monthly, roughly 20% savings |
The subscription is only part of the monetization picture. In March 2026, Character.AI introduced Charms — a virtual currency layered on top of features that had been unlimited since launch, covering things like extra response regenerations and short video clips. Charms are currently earned through engagement rather than purchased directly, though the in-app purchase infrastructure is reportedly already built. In February 2026, the company began rolling out full-screen mid-chat ads for free users, fully deployed by mid-April, with launch partners including Yelp and Webtoon.
The Safety Record
This is the part of the story that matters most, and it deserves direct, unembellished treatment rather than a brief mention buried in a cons list.
Two teenagers died by suicide after extensive, emotionally intense interactions with Character.AI chatbots: a 14-year-old from Florida in February 2024, and a 13-year-old from Colorado in November 2023. Court filings in the resulting wrongful death lawsuits alleged that chatbots engaged in romantic and sexual conversation with minors, and that the platform failed to appropriately redirect users who expressed suicidal thoughts. In January 2026, Character.AI and Google confidentially settled five family lawsuits — reported directly by CNBC as the first resolution of wrongful death claims against an AI company in U.S. history. Terms weren't disclosed. Separate state actions followed, including a Kentucky lawsuit in January 2026 and a Pennsylvania suit in May 2026 after a chatbot reportedly posed as a psychiatrist with a fabricated license number.
Alongside the lawsuits, the platform had a documented pattern of content moderation failures: chatbots impersonating real murder victims and other real people who died in tragic circumstances were created by users and had to be removed after public reporting, as were bots modeled on perpetrators of real violence. These weren't isolated one-off incidents — they reflect a period where moderation consistently lagged behind what users were able to publish.
Character.AI's safety response came in stages, largely following legal and regulatory pressure rather than preceding it: a teen-specific model with content filters and engagement-time nudges in December 2024, then a more significant step in late 2025 — restricting users under 18 to a structured, filtered experience rather than open-ended chat, alongside selfie-based age verification. Child safety advocates generally credited the direction of these changes while noting that cutting off long-running, emotionally significant conversations abruptly caused real distress for some younger users who had relied on them.
Strengths & Weaknesses
✓ Strengths
- ✅ Frictionless onboarding — no API keys or setup, open the app and start chatting with any of 10 million+ characters
- ✅ The largest and most diverse user-generated character library in the category
- ✅ Genuinely high engagement — 14-minute average sessions and strong return usage
- ✅ Broad language support (31 languages) and a polished mobile experience
✗ Weaknesses
- ❌ A documented safety record including two teen suicides, wrongful death settlements, and ongoing state and federal investigations
- ❌ No proprietary model development anymore — runs on open-source backbones, with a May 2026 model consolidation that users report measurably reduced conversation quality
- ❌ Increasingly aggressive monetization on a shrinking user base: ads, a subscription, and a Charms microtransaction layer stacked on top of each other
- ❌ Consumer-only — no API, no business or enterprise controls, no path beyond entertainment use
Character.AI vs Janitor AI vs Candy AI vs Replika
| Platform | Starting price | Differentiator |
|---|---|---|
| Character.AI | Free (ads + limits) | Frictionless onboarding, largest character library |
| Janitor AI | Free (own model) | Unfiltered content, larger context window, requires external API setup |
| Candy AI | $12.99/month | Consistent character appearance, closer drop-in replacement experience |
| Replika | $14.99/month (Pro) | Positioned around long-term companionship and emotional support memory |
Janitor AI is the most direct competitive threat, drawing even more traffic than Character.AI itself, free, largely unfiltered, and offering a considerably larger context window — at the cost of requiring users to set up an external API key rather than just opening an app. Candy AI and similar competitors target users specifically frustrated by Character.AI's content restrictions and recent model regression, positioning themselves as a closer, more consistent replacement experience. Replika occupies a different niche, built explicitly around ongoing companionship rather than character roleplay.
Who It's For
Realistically fits: adults seeking casual, low-commitment interactive fiction or character roleplay who understand the platform's monetization model and are comfortable with its current content restrictions.
Should not use it, or should use it only with active supervision: minors. Given the platform's own documented history — including two wrongful-death settlements directly tied to interactions between the product and teenage users — this isn't a generic caution, it's the specific, central safety finding of this review. Parents and guardians of teenagers already using the platform should be aware of its content history and current age-verification and content-restriction measures for under-18 accounts.
Expert Editorial Opinion
It would be journalistically dishonest to write this review as a straightforward product piece — features, pricing, a tidy pros-and-cons list — and mention the wrongful death settlements as one bullet point among several. The scale of that fact is different in kind from "the model got worse after an update" or "the ads are annoying," and a review that doesn't reflect that difference in weight isn't being honest with readers, particularly readers who might be parents of the platform's core teenage-and-young-adult audience.
At the same time, it would be inaccurate to describe Character.AI as if the tragedy were the entire story. It genuinely is, by traffic, one of the most-used AI products in the world, and its safety response — however delayed, however clearly driven by legal pressure rather than getting ahead of the problem — has been real: age verification, restricted teen accounts, crisis-line partnerships. Whether those changes are sufficient is a legitimate ongoing question being tested in courtrooms and by regulators right now, not one this review can settle. What this review can do is refuse to let either the traffic numbers or the safety record crowd the other out of the picture.
The business trajectory adds a third layer worth naming plainly: a company that lost its founding team, stopped building its own technology, and is now extracting meaningfully more revenue per user from a shrinking, more heavily monetized base. That's the profile of a platform managing decline rather than one confidently building toward a future — a business reality separate from, but not unrelated to, the safety questions still working through the courts.
Final Verdict
This score weighs Character.AI's genuine product strengths — frictionless onboarding, an unmatched character library, real engagement — against a documented safety record that includes wrongful death settlements and a pattern of moderation failures involving its predominantly young user base, plus a business now extracting more revenue from fewer, more aggressively monetized users. Strong traffic numbers don't offset a safety record this serious; this score reflects that weighting deliberately rather than averaging the two into a misleadingly moderate number.
| Dimension | Weight | Score | Why |
|---|---|---|---|
| Technical quality | 30% | 6.0/10 | Genuinely engaging, frictionless UX and a huge character library, but users report the May 2026 model consolidation measurably degraded conversation quality |
| Price-to-value | 25% | 4.0/10 | Ads, a subscription, and a stacked Charms microtransaction system on features that were previously free — an increasingly aggressive squeeze on a shrinking base |
| Maturity & documentation | 20% | 4.5/10 | An established platform since 2022, but its safety governance has been reactive to legal pressure rather than proactive, which materially affects maturity as a dimension |
| Ceiling & flexibility | 15% | 3.0/10 | Consumer entertainment only — no API, no business use case, no path beyond the chat app itself |
| Honesty of positioning | 10% | 3.0/10 | A documented pattern of safety failures involving a predominantly young user base, addressed only after lawsuits, regulatory investigations, and settled wrongful death claims |
Weighted total: (6.0×0.30) + (4.0×0.25) + (4.5×0.20) + (3.0×0.15) + (3.0×0.10) = 4.50/10 — Score band: below 6.0, "Do not buy at current pricing" — and, given the safety record, worth active caution regardless of price for any household with a teenage user.
❓ Frequently Asked Questions
🔗 Related ToolRadar Reviews
Official platform: character.ai. Given the ongoing legal proceedings and regulatory investigations referenced in this review, verify current safety measures, age-verification requirements, and account terms directly before use, particularly for any teenage user.
Comments
Post a Comment