When it comes to using generative AI in the workplace, adoption skews highest among currently employed college graduates, big earners, and those with full-time jobs, according to a new analysis from New York Federal Reserve researchers.
That raises questions about whether artificial intelligence might worsen labor market inequities — especially given that 62% of workers believe AI will lead to higher unemployment, the analysis found. Workers who earn less than $50,000 have an AI adoption rate of 15.9%, for example, while those who earn over $200,000 a year have an adoption rate of 66.3%.
Still, workers who earn the most also reported the firmest belief that AI will cause higher unemployment, at 69.4%.
Right now, AI’s impact on employment appears to be minimal, though some larger companies see workforce reductions in the year ahead, according to a separate analysis from the Federal Reserve Bank of Atlanta and Duke University’s Fuqua School of Business.
Though much has been made about the turmoil the tech could cause — or is already causing — among computer programmers and other white-collar workers, lower-wage workers are also fearful of losing their jobs. About 20% of the 30.6 million workers considered highly exposed to AI live in low-income households, according to an analysis from the San Francisco Fed.
Meanwhile, younger workers and others without college degrees want access to training, the New York Fed found. Only about 16% of employed survey respondents said their employer offered AI training, despite around 38% finding such upskilling important, largely because they hope it will make their jobs easier. They’re even willing to pay for the perk, especially if they’re young, a minority, lacking a college degree, or early in their job tenure — all categories of workers who tend to make less money.
“Crucially, some of the workers who place the highest value on AI training, such as those without a college degree, are also those with the lowest rates of AI usage and the lowest share of access to employer-provided training in how to use AI tools,” the New York Fed researchers wrote. “Closing this gap may be essential to reaching the productivity gains from having generative AI tools in the workplace.”
Employees without access to AI training would consider giving up an average of 11.4% of their current salary to switch to an employer that offers it, though the New York Fed reported that distribution was “highly skewed,” with the majority unwilling to accept any pay change.
Workers who already have access to training, meanwhile, would, on average, require a 24.2% salary bump to switch to an employer that lacks it, though these results could be affected by workers’ general unwillingness to sacrifice any part of their existing benefits package.