How many monthly active users does Character.ai have, and what are the implications for user data privacy?

Character.ai has roughly 20 million monthly active users, which makes the stakes of any security failure enormous, as seen in documented breach timelines across the AI companion industry.

What are the risks of sharing personal information with AI companions, and how can users mitigate these risks?

Sharing personal information with AI companions can lead to unintended consequences, such as targeted advertising or data sharing with third parties, as noted by Jennifer King from the Stanford Institute for Human-Centered AI; users can mitigate these risks by avoiding the sharing of identifying information and using unique emails.

What is the importance of transparency in privacy policies for AI companion apps, and how can users evaluate these policies?

Transparency in privacy policies is crucial, as users deserve to know what data is collected, where it goes, and whether it trains the models; users can evaluate these policies by looking for plain-language answers to these questions, as well as clear opt-out mechanisms, such as those required by California's SB 243 .

How can users protect their privacy when using AI companion apps, and what features should they look for in a platform?

Users can protect their privacy by treating every chat as potentially leakable, avoiding the sharing of identifying information, and using unique emails; they should also look for features such as in-app reporting, hard-coded prohibited categories, and honest limits, rather than 'unrestricted' promises, as seen in platforms like FlirtAlly.

What are the potential consequences of integrating AI into secure messaging, and how can users ensure their data is protected?

Integrating AI into secure messaging can compromise end-to-end encryption, exposing data to potential leaks, as noted by the ACLU ; users can ensure their data is protected by choosing platforms that prioritize user privacy and security, such as those that use end-to-end encryption and secure data storage.

How can users evaluate the privacy practices of AI companion apps, and what red flags should they look out for?

Users can evaluate the privacy practices of AI companion apps by reading the privacy policy, looking for clear opt-out mechanisms, and checking for transparency in data collection and sharing practices; red flags include vague data practices, lack of transparency, and undisclosed data practices , as highlighted by the FTC's inquiry into AI chatbots.

Character AI private messages: how to stay in control

Your AI companion knows things you'd never tell a stranger — your fears, your relationships, your 2 a.m. confessions. The question is who else ends up knowing them.

Character.ai's scale makes this urgent: the platform has roughly 20 million monthly active users, each generating intimate conversation data around the clock. When that much sensitive material sits on company servers, the stakes of any security failure are enormous.

The risks aren't hypothetical. Documented breach timelines across the AI companion industry show that user conversations can and do leak — which is why security writers consistently advise treating every chat as potentially leakable and never sharing identifying details with a bot.

Transparency is the other half of the problem. Privacy analysts note that AI companion apps often fail to clearly explain what data they collect, how long they store it, and who it's shared with. Regulators have started pushing back: Replika's parent company, Luka, was fined €5 million by the Italian data protection authority in 2025 for GDPR violations, and California's SB 243 — effective January 2026 — now requires AI companion apps to disclose their data practices, with users able to sue for $1,000 or more per violation.

The technical architecture adds another layer of concern. According to the ACLU, integrating AI into secure messaging can compromise end-to-end encryption, exposing data to potential leaks that users assumed were impossible. And Jennifer King of the Stanford Institute for Human-Centered AI warns that sharing sensitive information with chatbots may lead to unintended consequences, including targeted advertising or data sharing with third parties.

So what can you actually control? A few habits go a long way:

Treat every chat as potentially leakable — assume a breach is possible before you type.
Never share identifying information: real names, addresses, workplaces, or financial details.
Use a unique email for companion apps so a breach can't cascade into your other accounts.
Read the privacy policy before getting attached — look for data retention and sharing terms.

Regulatory attention is growing — the FTC launched an inquiry into AI chatbots acting as companions in September 2025 — but rules take time. In the meantime, the burden of caution sits with you.

It's a burden platforms should share. That's why FlirtAlly builds its companion experience around user control — with in-app reporting, hard-coded prohibited categories, and honest limits rather than "unrestricted" promises — so staying in control isn't something you have to do alone.

Transparency and User Control

A privacy policy nobody reads is not consent — it's a shrug. As AI companion apps grow, the gap between what platforms collect and what users understand keeps widening, and regulators are starting to close it.

The stakes are real. Replika's parent company, Luka, was fined €5 million by the Italian data protection authority in 2025 for GDPR violations, according to a review of AI companion privacy risks. That fine sent a clear message: vague data practices now carry consequences, not just criticism.

Legislation is catching up too. California SB 243, effective January 2026, requires AI companion apps to disclose their data practices and gives users a private right of action to sue for $1,000 or more per violation. Transparency is no longer a courtesy — it's a legal obligation.

So what does genuine transparency look like? At minimum, users deserve plain-language answers to three questions: what data is collected, where it goes, and whether it trains the models. Jennifer King of the Stanford Institute for Human-Centered AI warns that sharing sensitive information with chatbots can lead to unintended consequences like targeted advertising or third-party data sharing — outcomes most users never knowingly agreed to.

Clear opt-out mechanisms matter just as much. Users who don't want their conversations used for AI training should be able to say no through an accessible, one-step control, not a buried settings menu. That's the standard the FTC has signaled it expects, having launched an inquiry into AI chatbots acting as companions in September 2025.

For users evaluating any platform, a few checks separate the transparent from the evasive:

Does the privacy policy state plainly what is logged, stored, and shared — without legalese buried on page nine?
Is there a real opt-out for training use, reachable in a couple of clicks?
Can you delete conversations and account data on demand?
Does the app disclose monitoring or review of private chats up front?

This is the lens we applied when building FlirtAlly: message controls and monitoring practices are explained directly to users rather than hidden in fine print, because staying in control starts with knowing what's happening to your data. With Character.ai alone serving roughly 20 million monthly active users, the platforms that treat transparency as a feature — not a liability — are the ones that will keep their users' trust.

Best Practices for User Privacy

The most effective privacy strategy for AI companion chats starts with a mindset shift: treat every conversation as potentially leakable. That advice comes from an analysis of the AI companion industry's data breach history, which shows that even popular platforms with millions of users have exposed private chats through security incidents and poor data handling.

The scale of the risk is real. Character.ai alone serves around 20 million monthly active users, and that growth has outpaced the industry's privacy safeguards. Regulators have taken notice — Italy's data protection authority fined Replika's parent company €5 million in 2025 for GDPR violations, and the FTC launched an inquiry into AI companion apps in late 2025 over how these services handle user data and safety.

So what can you actually do? Privacy researchers and security writers converge on a handful of practical habits:

Never share identifying details — your full name, address, workplace, or phone number — in chat with an AI companion.
Use a unique email address for your companion app account, separate from your primary inbox.
Read the privacy policy before you invest months of conversation, paying attention to data collection, storage, and sharing practices.
Look for clear opt-out mechanisms if you don't want your conversations used to train AI models.

The stakes go beyond a hypothetical breach. Jennifer King of the Stanford Institute for Human-Centered AI warns that sharing sensitive information with AI chatbots can lead to unintended consequences, including targeted advertising or data sharing with third parties. The ACLU raises a related structural concern: integrating AI into secure messaging can compromise end-to-end encryption, exposing data to potential leaks.

Regulation is starting to give users more leverage. California's SB 243, effective January 2026, requires AI companion apps to disclose their data practices — and gives users a private right of action to sue for $1,000 or more per violation. That means platforms now face real legal consequences for opaque handling of your private messages.

Platform design matters too. FlirtAlly, for example, keeps chat monitoring and message controls in the user's hands — PG/Flirt is the default experience, the age-verified Spicy side is strictly opt-in, and in-app reporting is built in rather than buried. When evaluating any companion platform, ask the same questions: can you control who sees what, can you report problems easily, and does the company disclose its data practices up front?

The bottom line: privacy in AI companion chats is a shared responsibility, but the user-side habits are yours to own. Assume chats are leakable, share nothing identifying, use a dedicated email, and favor platforms that are transparent by design. The industry is being pushed toward honesty by regulators and users alike — reward the platforms that get there first.

FlirtAlly's Solution for Secure AI Conversations

The privacy problems with AI companion apps aren't hypothetical. Replika's parent company, Luka, was hit with a €5 million fine by the Italian data protection authority in 2025 for GDPR violations, and California's SB 243 — effective January 2026 — now lets users sue AI companion apps for $1,000 or more per violation over undisclosed data practices. Against that backdrop, FlirtAlly is building its privacy posture around a simple idea: tell users the truth, and give them the controls.

That starts with honesty about what the app is and isn't. FlirtAlly is not an "unrestricted" platform, and it says so plainly. The default experience is PG/Flirt, and the 18+ "Spicy" side requires ID-based age verification plus a qualifying plan. Hard-coded prohibited categories, media scene screening, and in-app reporting round out the safety layer. In a market where vague privacy policies are a documented risk, that transparency is the point.

Memory control is the other half of the equation. Because companions remember everything — facts, conversations, emotional context — users need to know where that memory lives and who holds it. FlirtAlly keeps memory unlimited on every plan, including the free tier, so there's no incentive to pay for "better memory" as a privacy workaround. What you tell your companion stays in your relationship, not behind a tiered paywall that fragments your control.

The broader industry context makes this approach easier to appreciate. Privacy researchers warn that AI integration into secure messaging can compromise end-to-end encryption, exposing chat data to potential leaks, and experts recommend treating chats as potentially leakable — avoiding identifying information and using unique emails. Stanford's Jennifer King has noted that sharing sensitive information with chatbots can lead to unintended consequences like targeted advertising or third-party data sharing. The FTC opened an inquiry into AI companion chatbots in September 2025, signaling that regulatory scrutiny is only intensifying.

FlirtAlly's design choices map directly onto those concerns:

Default PG/Flirt mode, with age-verified Spicy access as a deliberate opt-in
In-app reporting so users can flag problems without leaving the platform
Media screening on every generated scene, not just text
Unlimited memory on all plans — no data-hoarding upsells

None of this means users should abandon basic precautions. The expert advice still applies: don't share identifying details with any AI companion, and keep your personal email separate from your companion life. But the platform you choose should meet you halfway. With Character.ai serving roughly 20 million monthly active users, the stakes for getting companion-app privacy right are enormous — and the companies that treat honest safety as a feature rather than a liability are the ones worth trusting with your conversations.

FlirtAlly is currently in pre-launch, with public registration closed and the product running on a waitlist. You can browse a read-only demo of a showcase account today, and join the waitlist to be first in when registration opens.

Character AI private messages — how to stay in control

Character AI private messages — how to stay in control

Key Facts

Transparency and User Control

Best Practices for User Privacy

FlirtAlly's Solution for Secure AI Conversations

Frequently Asked Questions

Control Is the New Trust

Character AI private messages — how to stay in control

Key Facts

Risks of Sharing in AI Companions

Transparency and User Control

Best Practices for User Privacy

FlirtAlly's Solution for Secure AI Conversations

Frequently Asked Questions

Control Is the New Trust