Artificial intelligence is rapidly becoming part of everyday life, from workplace tools to personal assistants. But a new study suggests that women, on average, view AI with greater caution than men, not because they oppose technology, but because they perceive higher risks.
The peer-reviewed research, published in PNAS Nexus, found that women are more likely to see artificial intelligence as risky, especially regarding job security, fairness, and long-term social impact. Researchers say this difference is closely tied to greater risk awareness and real-world exposure to potential AI disruptions.
The study was led by Beatrice Magistro from Northeastern University, along with her research team. They surveyed nearly 3,000 participants across the United States and Canada in November 2023, asking them whether they believed the risks of generative AI outweighed its benefits.
When respondents rated AI risk on a scale from one to ten, men gave an average score of 4.38, while women rated it higher at 4.87, about 11 percent more cautious. While the difference may appear small, researchers say it reflects a consistent and meaningful trend.
To better understand why this gap exists, the study also examined general attitudes toward risk. Participants faced decision-making scenarios similar to lottery choices, choosing between a guaranteed smaller reward or a higher-risk option with a bigger payoff. Women were more likely to choose the safer option, reinforcing the idea that higher risk aversion plays a role in their views on AI.
Another key factor was exposure to AI-related job disruption. Using education levels and job categories as indicators, researchers found that women are more likely to work in roles vulnerable to automation and technological change. This exposure may lead to stronger concerns about how AI could affect employment stability and career growth. The study also showed that women’s scepticism is not absolute. When researchers presented clear examples where AI adoption directly benefited workers, such as improving efficiency or reducing repetitive tasks, women supported AI use almost as strongly as men. This suggests that transparency and proven advantages can significantly shift perceptions.
According to the authors, these findings highlight the importance of developing AI policies that address gender-specific concerns. Without careful planning, they warn, artificial intelligence could unintentionally widen existing inequalities or fuel broader resistance to the technology.
The takeaway is clear: women’s caution around AI is rooted not in fear of innovation, but in practical concerns about fairness, security, and long-term impact. Addressing these issues openly may be key to building broader trust in the future of artificial intelligence.
