A recent AI-driven study published in PLOS One has uncovered evidence of gender bias in film criticism. The study, conducted by Jad Doughman and Wael Khreich of the American University of Beirut, revealed that reviews of movies with female-dominated casts contain significantly more sexist language than those featuring male-dominated casts.
The study analyzed 17,165 professional film reviews, combining data from the Movie Review Data dataset with metadata from the Open Movie Database. Using an AI-powered gender bias detection system, the researchers identified two well-documented forms of sexism: benevolent sexism, which reinforces patronizing stereotypes of women as needing protection, and hostile sexism, expressed through overt negativity and aggression.
The results were as follows:
- Reviews from the study show that female-led movies include 149% more hostile sexism and 44% more benevolent sexism than those of male-led films.
- Benevolent sexism is prominent in the Family and Music genres, while hostile sexism peaks in Romance films.
- Compared to men, female actors, directors, and writers faced a higher degree of both benevolent and hostile sexist criticism.
Beyond linguistic bias, the study also highlights gender distinctions in the industry. Of the movies analyzed, 72% of lead actors, 91% of directors, and 86% of writers were male, reinforcing the industry’s long-standing underrepresentation of women.
The Power of Film Criticism in Shaping Careers
The study shows that film reviews shape public opinion and influence career trajectories, box office performance, and even mental well-being. “Our AI-driven analysis shows that systemic bias in reviews creates significant barriers to women’s advancement in the film industry,” said Doughman to Beirut Today. He adds, “This affects not only their earnings but also their well-being.”
While previous studies have examined gender bias through box office numbers and ratings, Doughman and Khreich’s research highlights how language reinforces gender disparities. Their findings suggest that film critics have the power to promote a more equitable industry by challenging rooted biases in their writing.
Next Steps: Addressing Bias in Global Cinema
During a phone call with Jad, currently managing his business outside of Lebanon, he noted that the study acknowledges its limitations, particularly its emphasis on Western films and critics. Expanding the dataset to include reviews from non-Western cinema could provide a more comprehensive picture of how gender bias manifests across cultures. While the AI models were used to detect sexism in review manuscripts, we augmented our dataset with gender information of the cast via name-based gender predictions. This approach did not capture non-binary and gender-diverse identities—an area for future research.
As the conversation around gender equity in Hollywood and global cinema continues, this study serves as a critical reminder: language matters. How we discuss films—and who gets to make them—has a lasting impact on the industry’s future.
The Psychological Impact of Biased Representation
In a conversation with child psychologist Ola Khoder, she highlights how children develop strong group identities early in life, which shapes their perceptions of gender roles and societal biases. “When children repeatedly see female-led films receiving harsher criticism or being framed in patronizing ways, it reinforces deep-seated stereotypes about women’s capabilities and roles in society,” she explained.
Khoder emphasized that language plays a crucial role in shaping young minds. “If critics consistently portray female directors, writers, and actors as less competent or more emotional, children absorb these messages and internalize them as norms,” she added. This contributes to a cycle in which gender bias is perpetuated across generations, affecting career aspirations, self-esteem, and even the way young audiences interpret storytelling.
She also pointed out that biased reviews not only influence perceptions of women in film but can also extend to real-life gender expectations. “How we discuss female-led stories, whether through subtle dismissiveness or overt criticism, teaches children what is valued in society,” Khoder noted.
The study mentions that in the film industry, men have historically dominated key roles, and these facts influence storytelling, representation, and movie reviews. While past research has focused on rating disparities between male- and female-led films, this study quantifies linguistic biases in professional reviews. By analyzing a large dataset, researchers found that female-led films face significantly higher levels of hostile and benevolent sexism in critiques. The findings emphasize the need for more balanced representation in the industry.
About the Researchers
Jad Doughman is an AI researcher and engineer specializing in natural language processing. A graduate of the American University of Beirut and recipient of the CCT Computer Science Student of the Year Award, his work has been published at ACL and EMNLP, leading conferences in AI research.
Wael Khreich is an Assistant Professor of Machine Learning at AUB’s Olayan School of Business. His research focuses on ethical AI, machine learning, natural language processing, and business analytics, with a particular interest in AI-driven bias detection.
