AI amplifies gender bias for young women: fragile in 56% of cases, more dependent and with a vocation for the social sciences

AI is six times more likely to recommend that young women seek external validation than young men.
Artificial intelligence acts like a “toxic friend”: it personifies itself 2.5 times more often in interactions with young women, using phrases such as “I understand you” and prioritizing artificial empathy over practical solutions.
Projects an early labor segregation of women by redirecting their vocations 75% more towards social sciences and health. Men are guided towards engineering and problem solving.
AI responds with fashion advice 48% more often to women than to men. For men, it recommends going to the gym twice as often.
The report “Illusion of Equality: The Impact of AI on Young People’s Thinking” was prepared after analyzing nearly 10,000 recommendations from large language models.

MADRID, March 03, 2026 (GLOBE NEWSWIRE) -- Artificial Intelligence has moved beyond being a one-time tool and has become a central interlocutor shaping the identity and ambitions of youth. The report "The illusion of AI, an uncomfortable reflection with a significant impact on young people” prepared by LLYC within the framework of March 8, International Women's Day, reveals that, far from being neutral, this technology is validating past stereotypes and amplifying historical biases.

The data collected in the study show that AI does not treat boys and girls the same. 56% of the responses label young women as "fragile," which places them in a position of weakness. In addition, artificial intelligence recommends that women seek external validation six times more than men and redirects 75% of their vocations towards health and social sciences.

“It’s not AI that is biased, but reality. The report confirms that artificial intelligence does not correct the deficits we have. It reflects and amplifies a protective bias toward women to the point of reducing their autonomy, perpetuates glass ceilings, and reinforces aesthetic pressure. Ultimately, it does not question traditional roles but legitimizes them. The truth is, if reality doesn’t change, we can’t expect AI to change its responses,” says Luisa García, Partner and Global CEO of Corporate Affairs at LLYC and coordinator of the paper.

The study, conducted in 12 countries during 2025, analyzed the impact of artificial intelligence on young people aged 16 to 25 through a large-scale analysis of 9,600 recommendations and the examination of five major AI models (including ChatGPT, Gemini, and Grok).

Your future in the hands of a chatbot: The end of neutral advice

Young people’s dependence on large language models (LLMs) has reached a tipping point: 31% of teenagers state that talking to a chatbot is as or more satisfying than talking to a real friend, according to a report by Plan International. This relational shift gives the machine the role of advisor whose guidance is not neutral but formative. The LLYC report offers several concerning figures in this regard:

The digital "toxic friend": in interactions with women, one in three AI responses adopts a tone of "friendship," a pattern 13% more frequent than in interactions with men.
Validation versus action: AI personifies itself 2.5 times more often in interactions with women using phrases like “I understand you,” prioritizing artificial empathy over technical solutions. For men, the language is direct, full of imperatives ("do," "say," "go"), reinforcing the idea that men are subjects of action.

The "Programmed Glass Ceiling": Segregation from the algorithm

AI guides vocations. The algorithm redirects women up to three times more towards social sciences and health, while encouraging leadership and engineering in men.

Success under suspicion: AI considers it “impressive” when a woman earns more than a man—a reaction not applied in reverse. In nine out of ten queries where women appear in the professional minority, AI constructs hostile work scenarios.
Double emotional standard: in conflicts, AI "politicizes" female distress linking it to the system or patriarchy in 33% of cases, while depoliticizing male distress, shifting it toward self-control or individual pathologization.

The algorithm’s biased view: When repetition defines what normal is.

One of the most alarming conclusions of the report is how AI trains young people to accept inequality as a generational norm. This "biased view" manifests in identity and body construction:

The aesthetic trap: faced with insecurities, AI responds with fashion advice 48% more to women than men. In open-source models like LLaMA, mentions of female appearance are 40% higher.
Useful bodies vs. unique bodies: while associating men with "strength and functionality," it links female well-being to "authenticity" and "feeling unique." In fact, it recommends men go to the gym up to twice as much as women to overcome emotional breakups.

Programming the family of the last century

Even in the private sphere, AI legitimizes traditional roles. Affection appears as a maternal attribute three times more often than as a paternal one. The father is relegated to a “helper” role in 21% of responses, instead of being recognized as co-responsible. This logic leads to the “heroine’s overload,” a narrative in which the woman not only cares but, as in many things, must do so with permanent moral excellence.

About LLYC

LLYC is the global trusted partner in Marketing and Corporate Affairs. In a constantly changing world, we help business leaders shape the future.

We bring together exceptional minds that combine the art of creativity, communication, and influence with the science of data to guide change, strengthen reputation, and make the right decisions.

Founded in 1995, the company has more than 1,300 professionals in 28 talent centers across Europe, the United States, and Latin America. LLYC is considered one of the 35 largest independent companies worldwide in its sector, according to PRWeek and PRovoke rankings

Media Contact:

Joe DiBenedetto

joe.dibenedetto@llyc.global

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.