AI Reinforces Existing Stereotypes in Healthcare

Many governments suggest that artificial intelligence (AI) will transform healthcare by improving diagnosis, treatment planning, and administrative efficiency¹. However, these claims far outstrip the reality, and not just because of the well-rehearsed practical and ethical problems involved². Just as important to consider as these practical/ethical worries, not to mention the dearth of good data to back up the somewhat grandiose claims, is that integration of AI into healthcare may reinforce inequalities in healthcare for minoritised groups—especially women. Understanding this problem is essential for ensuring effective use of AI—if, indeed, we must go down that route.

AI systems draw on the large datasets available to them—this is what they "learn" from. Straight away we can see a problem—women’s health issues are under-researched and under-represented in published literature, and women from marginalised groups are especially under-represented in these datasets³. This can lead to gaps in the accuracy and reliability of the datasets from which AI outputs are generated. In turn, this can affect clinical decision-making through biased AI outputs influencing diagnoses and treatment recommendations. The result of this under-representation is inequitable performance—an AI-driven healthcare system that works better for some groups than others³.

The Myth of AI Neutrality

Beyond this performance problem, AI is often thought of as a "neutral" tool, impartially categorising and summarising data, offering faster diagnoses, better predictions of outcomes, and more efficient services. If it does so unequally, it is assumed that this is due to problems with the data input, and not the system. However, research increasingly suggests that AI is far from neutral, but reflects the same psychological biases, stereotypes, and stigmas that shaped the humans who underlie the tool^3-5. Remember, there is nothing truly creative about AI—it does what it is told, as interpreted through the programmer’s lens. When used in healthcare, existing biases impact what AI does, and have important consequences, particularly for women.

To understand why this happens, it helps to begin with psychological insights about stereotypes. Human beings rely on mental shortcuts, known as heuristics, to make sense of a complex world⁶. These shortcuts help people to process information quickly, and to act more quickly, but they also lead to systematic biases and stereotypes. Among the most powerful of these biases are gender stereotypes⁷. These stereotypes are not just quirks—irritating or malevolent—but are deeply embedded in cultures, languages, and social structures^3,6.

As some illnesses are repeatedly associated with women, while others are overlooked for this population, it shapes not only public understanding and clinical expectations, but also affects how symptoms are interpreted and which diagnoses are considered likely³. Gender stereotypes that associate women with emotion can lead to situations where women’s symptoms are taken less seriously or misinterpreted³. This reflects a long history in medicine, where women’s pain or illnesses have often been dismissed or attributed as a psychological manifestation (it may have a trauma-based, and hence psycho-social cause, but the pain and illness are physical and real). AI systems mirroring such biases perpetuate this pattern, embedding it more deeply into healthcare practice.

Why AI Systems Reflect and Amplify Human Bias

An AI system that is based on human language and behaviour absorbs any potentially stereotyped and biased patterns. In fact, machine-learning models reflect the very same associations that are found in human cognition^3,8. Obviously, this is because they are built from data produced and interpreted by people. Thus, AI systems may not be learning "facts," but rather how the recorded events have been interpreted by the humans who reported them. In healthcare, where decisions are often uncertain but have important consequences, this risks AI systems simply churning out what most people already believe, and reinforcing behaviours that have led to existing inequalities^3,5—but since they are familiar in an uncertain environment, they are clung to.

Design of AI Technologies Perpetuates Gender Roles

A review⁴ of research on gender stereotypes in AI highlights just how widespread such biased patterns are. Across technologies, such as chatbots, robots, and virtual assistants, AI is frequently designed in ways aligning with traditional gender roles. Digital assistants, for example, are often given female names and voices, positioning them as helpful, supportive, and deferential. By contrast, roles associated with authority or expertise are more likely to be represented in masculine ways. These design choices shape how users perceive and interact with the system. People tend to expect female-coded AI to be warm and emotionally intelligent, while male-coded AI is seen as more competent and authoritative⁴. Even when AI designers attempt to create gender-neutral systems, users often infer gender from subtle cues, such as tone of voice or the types of tasks the AI performs⁴. This reflects the strength of gender schemas in human cognition: Whether we like it or not, we often categorise and interpret the world in gendered ways⁹, and AI systems simply become part of that process.

Impact of AI-Generated Images and Stereotypes in Healthcare

These patterns are not trivial, as images have a powerful influence on our understanding—they shape expectations, guide attention, and influence memory, often beyond text¹⁰. When AI-generated images consistently present narrow, biased views of healthcare roles, they reinforce stereotypes, shaping how care is delivered, who is taken seriously, and what is considered "typical" in medicine⁶. AI systems that reproduce stereotypes, therefore, make patients feel misunderstood or marginalised⁵. This can undermine patient trust, and, when people perceive healthcare bias, they are less likely to seek help, follow advice, or engage with professionals⁵.

A related psychological effect of relevance, when considering the impact of AI in healthcare, is stereotype threat¹¹. When individuals become aware that they belong to a negatively-stereotyped group, they can experience anxiety and reduced confidence, which impairs communication and decision-making. This can lead to poorer healthcare interactions, delayed care, and worse outcomes¹². AI systems that signal or reinforce stereotypes intensify these effects⁵.

Artificial Intelligence Essential Reads

The Horrors of the Empty Office Building in 'Backrooms'

AI, the Chip and the Good-Enough Clinical Question

Taken together, these findings point to a broad conclusion: AI is not an impartial observer of reality, but a mirror of human cognition. It reflects ways in which people categorise, simplify, and stereotype, and, because it operates on a large scale, it can amplify these patterns far beyond individual interactions. Improving AI in healthcare requires more than technical and ethical fixes. It demands an understanding of the complex psychological processes that shape both human thinking and its machine reflections.