Rizz's 7.5M Downloads: A Wake-Up Call for Dating Apps Ignoring AI

Match Group and Bumble have spent years optimising their apps to keep conversations flowing. Turns out, 7.5 million people have decided they need an AI to handle that bit for them. Rizz, an AI-powered dating assistant that analyses screenshots from dating apps and generates conversation suggestions in real time, has reached a scale that represents not a niche curiosity but a genuine behaviour shift in how people approach digital romance.

The app positions itself as a 'wingman' that helps users craft responses when they're stuck—upload a screenshot of your chat, get AI-generated reply options, copy and paste. The company claims its technology 'enhances rather than replaces connection', a framing that deserves considerable scepticism given the product literally automates the most human part of digital dating: the actual conversation. What makes this worth watching isn't the technology itself—large language models generating flirty banter is hardly groundbreaking engineering—but the scale of adoption amongst the dating app user base.

Create a free account

Unlock unlimited access and get the weekly briefing delivered to your inbox.

Seven and a half million downloads represents a meaningful slice of the dating app user base actively choosing to outsource their personality to an algorithm.

Whether this represents genuine user need or learned helplessness in digital communication is a question with serious implications for how operators think about product development and user success metrics. Dating apps have taught users that romance is an optimisation problem, but they assumed conversation quality would remain human territory. That assumption now looks vulnerable.

The outsourced authenticity economy

The 7.5 million figure—whilst unverified by independent sources—sits within a credible range given the broader AI assistant market. For context, ChatGPT reached 100 million users in two months. Dating-specific AI tools are proliferating: apps that write your bio, optimise your photos, generate opening lines, even conduct preliminary screening chats.

Rizz's specific pitch is real-time conversation rescue, pitched at the moment of friction when someone doesn't know what to say next. Roman Khaves, the app's developer, frames this as assistive technology rather than deception. The comparison to calculators solving maths problems is one he's reportedly made—a tool that handles the mechanical labour whilst the human provides the intent and judgment.

But the analogy breaks down quickly. Calculator-assisted maths doesn't claim to reveal your personality. An AI drafting your flirty messages absolutely does misrepresent who's on the other end of the conversation. The privacy angle here warrants closer examination than Rizz's promotional materials provide.

Users are uploading screenshots containing conversations, profile information, and potentially identifying details about matches. The company's assurances about data handling are just that—assurances. Dating app operators have compliance teams dedicated to GDPR, the Online Safety Act, and platform-specific privacy frameworks. Third-party screenshot tools operate in murkier regulatory territory.

What the platforms aren't saying

Dating apps have remained conspicuously quiet about the rise of AI assistants parasitically attached to their ecosystems. There's no technical way to detect whether a message was human-written or AI-generated. Match Group's portfolio hasn't publicly addressed whether AI-assisted chatting violates terms of service. Bumble's brand positioning around women making the first move becomes somewhat hollow if that first move was written by an algorithm.

The silence likely reflects an uncomfortable reality: the platforms can't stop this, and acknowledging it openly risks legitimising behaviour they'd rather pretend doesn't exist at scale. Telling 7.5 million people they're 'cheating' at dating would be tone-deaf at best. Ignoring it means letting a third party effectively mediate the core value proposition—human connection—that dating apps are supposed to facilitate.

If millions of users clearly want AI communication assistance, should platforms build it natively rather than cede that functionality to external tools?

From a competitive standpoint, this creates an odd dynamic. If millions of users clearly want AI communication assistance, should platforms build it natively rather than cede that functionality to external tools? Doing so would at least keep the data in-house and allow for integration with match algorithms and safety features. But it would also represent an admission that modern dating app users need algorithmic help not just to find matches, but to talk to them—a fairly damning indictment of either the user base's communication skills or the platforms' success at fostering genuine interaction.

The regulatory vacuum

Khaves has mentioned government scrutiny in recent coverage, though specifics remain vague. The regulatory framework for AI in intimate contexts is still forming. The EU's AI Act classifies some emotion recognition and biometric systems as high-risk, but conversational AI assistants likely fall outside current scoping. The OSA focuses on platform duties of care, not third-party tools that analyse screenshots.

What regulators should be watching—and dating app compliance teams ought to be tracking—is the intersection of AI assistance and trust and safety. If AI tools are crafting messages designed to maximise engagement or attraction, are they also capable of recognising and flagging manipulation, love bombing, or grooming behaviours? The incentive structures are misaligned. Rizz succeeds when users get responses and maintain conversations. Safety-conscious moderation would sometimes mean suggesting users disengage.

The other regulatory risk is consent. Does the person on the receiving end have a right to know they're chatting with human-plus-AI rather than just human? Dating apps require real identities and ban impersonation. An AI writing your personality arguably constitutes a form of identity misrepresentation, even if the face in the photos is yours.

What adoption signals

The charitable interpretation of Rizz's 7.5 million downloads is that digital communication is a learned skill, unevenly distributed, and many people simply want help crafting thoughtful responses rather than relying on 'hey' or emoji. The less charitable read is that a generation raised on algorithmic feeds and parasocial relationships now lacks the confidence—or perhaps the interest—in unmediated human conversation, even when pursuing romantic connection. Both can be true.

The dating industry has spent fifteen years teaching people that romance is a swipe-and-optimise problem. Conversation quality was always assumed to be the user's responsibility, the part that automation couldn't touch. Rizz's success as the 5th most downloaded dating app suggests that assumption was wrong, or at least that millions of users are willing to test whether AI can handle it.

For operators, this trend presents a strategic fork. Build AI assistance into the product and own the entire user journey, risking brand damage and authenticity concerns. Or maintain the current position—that conversation is human territory—whilst watching third parties capture value and user attention at a critical conversion point. Neither option is comfortable, but pretending 7.5 million people aren't already automating their chat is no longer viable.