Educational technology has been linked to falling test scores in Utah and other countries by neuroscientist and former teacher Jared Cooney Horvath, who says the widespread use of computers and tablets in classrooms may be weakening the cognitive skills schools are meant to develop.
Horvath said the decline in Utah’s reading and maths results began around the introduction of the state’s first computer-adaptive assessment, the Student Assessment of Growth and Excellence (SAGE), in 2014.
Before then, computers were present in schools but played a limited role, he said. The new testing system required schools to build the digital infrastructure needed to administer assessments electronically.
Horvath, author of The Digital Delusion: How Classroom Technology Harms Our Kids’ Learning—and How to Help Them Thrive Again, said Utah’s experience formed part of a broader international pattern. He told a US Senate committee that falling standardised test scores had coincided with the rapid expansion of classroom technology.
He also pointed to results from the Programme for International Student Assessment, which tests 15-year-olds around the world. The data, he said, showed an association between declining scores and the amount of time students spent using computers.
Horvath argued that the issue extended beyond examination results, claiming that today’s generation was the first in modern history not to outperform its parents on standardised assessments. He described this as evidence that Gen Z was less cognitively capable than previous generations.
The expansion of educational technology
Technology was introduced to classrooms with the aim of improving learning and allowing lessons to be adapted to individual pupils. Horvath said the education technology industry instead promoted the idea that schools were fundamentally broken and that computers could provide the solution.
The expansion gathered pace in the United States after Maine introduced a statewide laptop programme in 2002. The initiative initially distributed 17,000 Apple laptops to seventh-year pupils across 243 schools; by 2016, 66,000 students in the state had laptops or tablets.
By 2024, US schools had spent more than 30 billion dollars on classroom screens and related equipment, with districts securing deals for discounted devices. Google Chromebooks, which offered low-cost hardware and access to free Google applications, accounted for more than half of digital devices supplied to schools by 2017.
Horvath claimed that some technology companies benefited commercially from the expansion, including Google after the uncertain launch of its Chromebook. Google did not respond to a request for comment included in the source material.
He said the underlying educational weakness was what he called the “transfer problem”: pupils could perform tasks while using a particular device or programme, but struggled to apply the knowledge away from it.
The problem, he argued, was not new. Teaching machines developed in the 1920s and 1950s allowed pupils to work through questions at their own pace, but failed to convince educators that they improved learning. Their critics found that students could become proficient at operating the machine without mastering the subject itself.
Artificial intelligence and the classroom
Horvath said the same concern had re-emerged with artificial intelligence. A Pew Research Centre survey cited in the source material found that more than half of US teenagers used AI for schoolwork, while a Brookings report suggested some students were using it to cheat rather than learn.
He argued that meaningful learning required pupils to struggle with problems and develop their own reasoning. Experts could use AI effectively because they understood a subject well enough to assess and correct its answers; students, he said, risked learning dependence instead.
“The tools experts use to make their lives easier are not the tools children should use to learn how to become experts,” Horvath said. “When you use offloading tools that experts use to make their lives easier as a novice, as a student, you don’t learn the skill. You simply learn dependency.”
Some politicians have responded by seeking to restrict access. New York City mayor Zohran Mamdani announced a one-year moratorium on student-facing generative AI tools such as chatbots.
Horvath said a complete ban was not necessarily required. Schools could teach pupils about AI while continuing to build the underlying knowledge that would allow them to use it responsibly.
“If you really want kids to be good at AI, continue to teach them stuff. Teach them math, teach them literacy, teach them numeracy, give them a general education,” he said.
