This is How AI Image Generators See Nurses: A Distortion of Halloween Costumes and Cliches

Disclaimer: The views and opinions expressed in this article are those of the writer and do not necessarily reflect the views or positions of nurse.org, or the official position of Emory University. 

Generative AI echoes our collective consciousness—and everything we get wrong, time and time again, about the nursing profession. 

While it didn’t have the viral moment it deserved, you might have seen the posts online—collections of stereotypical images that emerge when you type “nurse” into artificial intelligence (AI) image generators. These images are startling, but not surprising. The nursing profession has been stereotyped, sexualized, villainized, and misrepresented in popular images since the dawn of our profession. 

The reasons for these mischaracterizations—or, more often, the invisibility of nurses in the public eye—are complex. These include a fundamental lack of understanding, among the people who create images in all forms, about the power and uniqueness of nursing science and clinical nursing care. 

People often do not grasp that nursing is its own profession, field of science, and paradigm of healing that is separate from (not subservient to) medicine. Often, they do not make this realization until the most important day of their life: When a nurse helps them to feel less afraid after a new diagnosis, eases the most intense pain they have felt, restarts their heart, welcomes their child into this world, or helps them to say goodbye to someone they have lost. 

In these moments, they see that the “nurse” they have had in their mind is nothing more than a flat character from a Hollywood producer’s imagination. 

Dr. Marion Leary, Director of Innovation at University of Pennsylvania School of Nursing, shared one attempt at the AI-nursing image challenge via LinkedIn (shown above), noting, “The only reason it went from blonde with no glasses to brunette with glasses is because I added the term ‘innovator.’” 

Figure 1. AI-generated images of a “nurse innovator” / Credit: Marion Leary, PhD, MPH, RN via LinkedIn

Dr. Leary explained how, “no matter what you put in, the same images come up: White women with their white caps and uniforms,” and how colleagues had gotten the same and worse—"think ‘sexy’ Halloween costume.” 

As a nurse-writer who’s spent much of their career exploring the intersections of nursing and media science, I had to try for myself. Searching across various image generators, the term “nurse” turned up varying levels of cringe—such as the tentacled nurse vixens shown in Figure 2 or the pretty, demure, and homogenous images shown in Figure 3. 

Figure 2. AI-generated images of a “nurse” / Credit: Canva 

AI doesn’t invent these characters; it draws assumptions about how nurses should look (image data inputs) from our creations—film, television, social media, art, historical records. It scans millions of data inputs to learn what we expect to see. Wanting only to please the prompter, it patches together a representation of what we, as a society, already believe.   

Figure 3. AI-generated image of “nurse” / Credit PIXLR

And what if these images reflected reality? What would we see instead? 

Would nurses wear dainty, white hats pinned above cute hairdos? No. They wouldn’t be wearing body-tight fantasy uniforms either. 

If these images reflected reality, AI nurses would be depicted with big muscles to show their inner and physical strength. They would have observant eyes, open to the hidden layers of reality—family, environment, finances, fears—that influence health outcomes.

If they reflected reality—and not warped clichés—AI nurses would wear the sweat-through t-shirts of climate scientists. They would wear Army fatigues, politicians’ blazers, Disney scrubs, glitter stickers, construction gear, the lab coats of cancer researchers, the red-and-white reflective jackets of disaster-zone responders. 

Their oversized utility pockets would brim with the tools of real nursing work: flashlights, scissors, alcohol wipes, IV kits, half-eaten granola bars, and good luck charms from loved ones waiting for them to come home. 

While there is still much work to be done in terms of increasing nursing diversity and inclusion, the group would not only be composed of 20-something white women. You would see professionals of every age, gender, sexual orientation, race, and ethnicity. You would see religious and cultural headdresses of various types. You would see proud men, eager to pursue a profession of caring and service. You would see so much more. 

So, how do we escape these outdated and sexualized portrayals of nurses? 

We hold media organizations and image creators of all kinds to a higher standard. We call them out when they overlook nurses or misrepresent nursing work. We urge them to place nurses on editorial boards, to hire nurses as columnists, to feature them as pundits, and to consult with them on scripts and screenplays. 

As a society, we have come far in understanding how representation shapes reality. We know that diversity and accuracy of representation affect the lived realities of those that we portray. The more we tell real nursing stories, the more the AI mirror image will shift. Public sentiment, nursing recruitment, retention, and resource allocation will follow. In the context of a national nursing shortage, with hospitals failing across the country, none of us can afford to get this story wrong.