AI Literacy Has A Class Problem: Study Shows The Width Of The Gap

Wealthier, Better-Educated Americans Know More About AI, and That Gap May Make Inequality Harder to Close

Every time a chatbot answers a customer service question, an email filter catches spam, or a social media feed serves up a recommended video, artificial intelligence is at work. Most Americans encounter that invisible machinery dozens of times a day without realizing it. According to a large national survey, that blind spot is not shared equally. Americans with higher incomes and more education are significantly better at recognizing when AI is involved in everyday situations, while those with less tend to miss it entirely.

Being unaware of AI isn’t a neutral position. People who can’t identify it can’t question it, push back, or take advantage of what it offers. A new study published in Information, Communication & Society finds that disadvantage tracks closely with existing economic divides.

Researchers at Hong Kong Baptist University analyzed survey responses from more than 10,000 American adults and found that socioeconomic status predicted not just whether someone recognized AI in daily life, but why. Two factors helped explain the gap: wealthier and better-educated people were more likely to use AI tools, and they also reported feeling more familiar with AI, meaning they said they had heard or read more about it. Both of those pathways fed into higher awareness scores.

How Researchers Measured AI Awareness

To conduct the study, the team drew on Wave 119 of the Pew Research Center’s American Trends Panel, a nationally representative survey of randomly selected U.S. adults. Data were collected December 12 to 18, 2022, just as the public boom around generative AI was beginning, so the awareness landscape may have shifted somewhat since then. After low-quality responses were removed, 10,087 respondents remained. About 62% identified as White, nearly 17% as Hispanic, roughly 12% as African American, and about 5% as Asian. Median education was some college without a degree, and median household income fell between $60,000 and $70,000.

Before any AI questions were asked, every respondent received a plain-language definition: AI is designed to learn tasks that humans typically do, like recognizing speech or pictures. Then came six multiple-choice questions testing whether they could identify which devices or applications in everyday scenarios actually use AI. Scenarios covered online customer service, email, and online shopping. One correct answer, for example, was recognizing that a chatbot immediately answering customer questions is powered by AI. Each right answer earned a point, for a maximum of six. Across all respondents, the average score was 3.72 out of 6.

Researchers also measured how often people used AI on a five-point scale and how familiar they felt with it on a three-point scale, asking how much they had heard or read about the technology. Education was recorded across six levels and household income across nine brackets. Age, gender, ethnicity, and geographic location were accounted for as control variables.

Higher Education Predicted AI Awareness More Than Income Alone

After accounting for those demographic factors, both education and income were significantly tied to AI awareness. Education turned out to be a stronger predictor than income, consistent with earlier research on internet usage gaps. Better-educated respondents scored higher partly because they were more likely to use AI tools regularly, and partly because they reported greater familiarity with it. Together, education, income, usage, and familiarity accounted for about 36% of the variation in awareness scores across the population.

Early internet users grew more savvy the more time they spent online, and something similar appears to happen with AI. People who regularly interact with AI tools develop a sharper eye for spotting it elsewhere. Those from lower-income and less-educated backgrounds, who use AI less and report hearing less about it, miss out on both of those learning channels.

Why the Awareness Gap Could Widen Economic Divides

Recognizing AI, the authors argue, is a foundation for AI literacy, the broader ability to use, evaluate, and think critically about the technology. People who can’t identify when they’re interacting with AI face barriers to questioning whether a system is biased or understanding how their data is being used. In areas where AI may influence consequential decisions, from job screening to lending, that blind spot carries real weight. AI systems can reflect and sometimes reinforce biases baked into their training data, and people unaware of AI’s role are less equipped to recognize or challenge those outcomes.

Men scored higher than women on AI usage, familiarity, and awareness, a finding the researchers attribute partly to cultural associations between technology and male identity. Younger respondents outscored older ones, consistent with younger people’s heavier engagement with AI-embedded platforms like social media and smartphones.

Addressing the gap, the authors argue, requires more than making AI tools available. Familiarity turned out to be a stronger predictor of awareness than usage, meaning exposure to information about AI through media, community programs, or school curricula may matter as much as hands-on access. Targeted outreach for lower-income and less-educated populations, using plain language and relevant examples, is where the effort needs to start.

AI awareness and economic advantage now fall along the same fault lines that have long divided American society. Whether that gap narrows may depend on who gets pulled into the conversation first.

---