First there were Instagram filters. Then, the photos generated by AI. Then, the “bios” written by ChatGPT. Today some users directly delegate the task of responding to messages to an algorithm. Artificial intelligence was introduced into dating apps almost imperceptibly and the result is an experience that is increasingly more polished, more strategic and, ultimately, less human.
The trap of the perfect profile
Given this context, Happn, the dating app based on real-life encountersdecided to set limits. It updated its Charter of Trust to explicitly prohibit three practices that have become increasingly common: creating profiles with AI-generated photos or descriptions, using chatbots to compose messages, and simulating false identities. Furthermore, he added a functionality that allows reporting suspicious profiles.
But it does not completely renounce artificial intelligence. Its Perfect Date feature uses generative AI to suggest places for a date based on interests. The difference, according to the company, is one of intention. “Artificial intelligence should be a support tool. It can help break the ice or suggest ideas, but it should never speak on behalf of people or replace the spontaneity of a real connection,” says Karima Ben Abdelmalek, CEO of the company.
The movement is not isolated. Bumble, one of the largest dating platforms in the world, announced something more drastic: it is going to eliminate the “swipe”, the gesture of swiping profiles. The decision comes at a critical moment for the company that lost about 21% of its paid users in the last year and is betting on a total renewal of its experience. Its CEO, Whitney Wolfe Herd, promised “something revolutionary for the category.” Everything indicates that the bet is strong on artificial intelligence: the company is already developing Bee, an AI-based dating assistant that, through a private conversation, seeks to know the valuesinterests and emotional expectations of the user to suggest compatible profiles.
invisible intermediary
Bumble knows how long we look at a photo before swiping. It knows what time we respond, how often, who we ignore. And with that information you can decide which profiles you show and, above all, which ones we will never see. “The algorithm can make you lose the love of your life,” warns Martín Pablo Villar, specialist in responsible automation and author of “Algorithmic privacy” (Hammurabi Publishing).
For him, the key is to distinguish between assisting a decision and replacing it. Having an app suggest you ten compatible profiles and choose one is one thing, but silently discarding hundreds of people before they reach the screen is quite another. “In the first case, technology works as a tool, in the second it begins to act as an intermediary that conditions our choices without us being aware of it,” he explains.
The problem deepens when it comes to the rules that the platforms themselves impose on themselves. Happn’s Trust Charter, for example, prohibits significant AI photo modificationsbut clarifies that “small tweaks” are acceptable. And what is a small touch-up? The company does not define it. What if someone with a disability uses a writing assistant to compose their messages? Is that cheating or is it accessibility? “These are issues that exceed the capacity of a company to resolve them unilaterally,” says Villar. Internal policies are necessary, he concedes, but insufficient. Without clear regulatory frameworks, each platform defines its own rules (and users are subject to standards that can be very different depending on the app).
In Latin America, the regulatory discussion on AI and digital platforms is advancing little and slowly. Argentina had a National Artificial Intelligence Plan in 2019 that was never executed and was abandoned by subsequent administrations. Europe, on the other hand, has been building a comprehensive approach for years: GDPR (General Data Protection Regulation) incorporates rights linked to automated profiling and algorithmic decision makingand more recent regulations such as the AI Regulation and the Digital Services Act seek to cover the specific risks of digital platforms. “The question is no longer whether AI is going to intervene in our personal relationships. That is already happening. The real question is who defines the rules of that intervention: whether technology companies alone, or democratic societies through institutions capable of establishing limits,” says Villar. For now, the answer is mostly the first option.
The last bastion
The other problem with this view is that desire does not function as a “checklist”. No matter how much an algorithm processes values, interests and emotional expectations, there is no system capable of predicting who someone will want to stay with. “Desire is something more unconscious. That is why there is no algorithm that can handle love”says Sofía Calvo, psychologist. For her, delegating to AI the choice of who to connect with is part of a broader and worrying process: the replacement of skills that should remain on the human side. Understand that seduction is not a minor procedure, because frustration, limits, and communication come into play. If technology replaces it, in the future we will not be able to interact without AI involved.
However, Calvo goes against the grain of technological panic. The frustration that comes with coming face to face with reality after weeks chatting with a profile that is too perfect is not necessarily a problem, but a sign that something human still resists. “I prefer that people get frustrated or feel deceived when they encounter reality. It is the last bastion to keep in mind that the perfect of the virtual has no place in the only place where we can create something human and therefore imperfect: the real world.”
You may also be interested
Image gallery
