groundbreaking research
May 1, 2026
People with lower socio-economic status are less likely to recognize and use AI tools, finds data on more than 10,000 US adults
Alamy: World Images Contributor
The widespread adoption of artificial intelligence (AI), especially in everyday “hidden” applications, is creating new and distinct forms of digital inequality.
That’s the warning from Hong Kong Baptist University communication researcher Professor Sai Wan and his colleagues, who analyzed data on more than 10,000 Americans’ interactions with AI in a paper published today in the journal Science. information, communications and technology.
The researchers’ analysis found that people with higher education and income levels were more aware of AI, more familiar with it, and more likely to use the fast-growing technology than those with lower socio-economic status (SES).
Researchers define AI awareness primarily as recognizing the use of technology in different situations. Familiarity, on the other hand, concerns people’s perceived knowledge about AI, regardless of their actual knowledge.
“Closing the AI awareness gap is essential, because if only those with higher incomes or higher education know about AI and its uses, it can further exacerbate social inequalities,” adds Professor Wang.
“This allows some groups to take advantage of advanced technology while leaving others behind.
“For example, job seekers who know that their employer is using AI in their selection process can better tailor their resumes, while job seekers who lack this awareness may unknowingly miss out on opportunities.”
Wang points out that while AI has the ability to empower individuals, it also comes with the risk of harm.
She explains: “Those with higher awareness may be better able to understand both the opportunities and risks of AI, such as recognizing and creating deepfakes, whereas those with lower awareness are more likely to be fooled or manipulated by these technologies.”
In their study, Wang and colleagues analyzed survey data about understanding and attitudes toward AI collected from 10,087 U.S. adults by the American Trends Panel, a nationally representative group conducted by the Pew Research Center in Washington, DC.
Respondents’ SES was assessed based on education level and household income, and the team found that the former was more closely associated with AI use.
Previous research suggests that wealthier and more educated people are more likely to be encouraged to utilize AI tools because they generally have more advanced digital skills, which in turn increases their confidence in using AI. These trends help explain why education and income emerged as important predictors of AI usage in the current study.
However, Wang said their study also revealed some unexpected findings. In other words, familiarity with AI is a stronger predictor of AI awareness than actual use of AI.
“In other words, simply feeling knowledgeable or informed about AI is more closely related to being aware of where it is and how it is being used, compared to personal use of AI technology,” Wang points out.
The explanation for this phenomenon may be that many common AI applications are so seamlessly integrated into the everyday apps and platforms of our digital lives that their additions are not obvious.
“For example, AI-driven recommendation systems on streaming platforms like Netflix and Spotify suggest content tailored to a person’s tastes,” Wang says. She continues: “However, many users are unaware that these are powered by AI and may think the recommendations are random or neutral.”
The new digital inequalities created by AI are thus different from the previous ones.
“Traditional digital inequality focuses on access, skills/use, and outcomes, all of which tend to assume that users consciously engage with technology,” Wang explains.
“However, AI is often integrated into everyday apps and platforms in ways that users don’t realize, and many people are unknowingly interacting with it through things like social media feeds and streaming recommendations.”
Therefore, increasing access to AI-powered technologies may not be enough to close this perception gap, the team explains.
Instead, researchers recommend an indirect approach to reducing this new digital inequality, especially by helping people from lower SES backgrounds understand important issues related to AI.
“This could include outreach campaigns and community workshops that use clear language and practical examples to make AI more understandable and relevant to low-SES communities,” Wang suggests.
The team hopes that resources will be made available to increase engagement on AI-related topics, address public concerns, and provide guidance on the ethical and responsible use of AI. Fundamental concepts of AI may also be incorporated into educational curricula.
The researchers added that AI literacy programs should include targeted guidance on how to identify “hidden” AI in everyday life and understand its basic functions.
“It is imperative that we work towards a more inclusive digital future, where technology empowers everyone and no group is further marginalized,” the researchers conclude in their paper.
The researchers cautioned that because it is US-centric, it is unclear how generalizable their findings are to other countries, where the level of adoption and awareness of AI may be different. For example, previous research has found that people in South Korea, China, and Finland have the highest awareness of AI, while the country with the lowest average awareness is the Netherlands.
With this initial study completed, the team is now looking to investigate how the digital divide manifests itself in the context of AI and what impact it has on society.
