Commentary: AI won't give American children the education they need
Published in Op Eds
Elected officials are finally waking up to the educational harms of mobile phones in public schools. As more districts ban them, the reports are highly encouraging — though hardly surprising, given the positive results we saw in New York City when we removed them from schools nearly 20 years ago.
Yet even as phone bans spread, elected officials and Silicon Valley executives are trying to open classrooms to a technology that could set students back even further than mobile phones have: artificial intelligence.
In early September, as millions of children were returning to school, technology executives and government officials gathered at the White House to discuss their vision for American education, one in which struggling students are guided by chatbot tutors, teachers are liberated from thinking and preparing thanks to automated lesson plans, and an army of teen AI innovators breeze through boot camps and certifications, ready to lead the workforce of tomorrow.
One could almost hear the sound of 76 trombones coming from the meeting, but tech companies that seek to sell hardware and software to American schools march to the beat of shareholders, not students — and if their utopian vision takes a dark turn, it is children and families who pay the price. If history is any guide, parents and educators should think twice before joining the big parade.
During the White House meeting, Google’s leadership touted AI as a means of transforming the learning process around each individual student and pledged access to the company’s AI tools for every high school in America. If that sounds familiar, it should. And it’s worth remembering what happened.
More than a decade has passed since Google and other tech companies started making big promises of personalized learning and improved academic performance by getting low-cost laptops into schools. That worked well for their sales departments, but not for taxpayers — it has cost billions of dollars — and especially not for students.
All the promised academic benefits of laptops in schools never materialized. Just the opposite: Student test scores have fallen to historic lows, as has college readiness. In some schools, students spend hours of valuable class time on YouTube and other social media channels. Teachers are too often reduced to IT monitors, frustrating them.
A growing body of research demonstrates the harms of excessive screen time for students, yet few districts have reviewed their policies on laptops — even as the prospect of AI in the classroom will mean more screen time, not less. Until more studies are done about AI’s impact on learning, greater caution is warranted.
Yet in April, the administration issued an executive order calling for children in the “earliest stages of their educational journey” to master AI. But AI is only a simple-to-use tool. The objective of public education should not be learning how to use any one tool. It should be mastering core concepts, developing critical thinking and reasoning skills, and imparting essential values, including those that AI struggles with, such as trustworthiness.
Nevertheless, Education Secretary Linda McMahon has said that any AI-related grant application would get “bonus points,” and she announced her plan to visit the flagship Alpha School in Texas — an experimental academy where traditional subjects like math and reading are taught by AI software and supervised by adult “guides.”
To be clear: I’m a big believer in AI. Our company has been investing in it for more than 15 years, and we have been increasingly incorporating it in ways that allow us to better serve our customers and allow our employees to perform higher-level work. But helping businesses serve customers and helping schools serve children are two very different challenges.
Early reports of AI’s impact on learning are not encouraging. One preliminary study published in June found that adults who use ChatGPT to write essays demonstrate weaker critical-thinking skills over time than participants who use search engines or no technology at all. It remains unclear how such tools will affect children’s brains over the long term, but the implications for writing and math skills development are cause for concern.
In time, AI-powered educational tools may prove useful in some contexts. A report published by the Organization for Economic Cooperation and Development found that an hour or less of computer-based learning was associated with higher math scores on an international assessment (effects that were reversed as screen time increased). Khan Academy’s AI bot Khanmigo also has shown promise — though the company acknowledges that “extended sessions can lead to repetitive responses or conversations that drift from educational purposes.”
As states develop and refine guidance to schools about AI implementation, they should require education officials to evaluate independent research before greenlighting AI products. Absent comprehensive data, AI use should be monitored, limited and geared toward older students.
Such deliberate planning isn’t unrealistic: Already, schools seeking to reduce screen time — by locking away phones and laptops and reopening mothballed computer labs — have reported increased engagement. Others that have designated “tech-free” school days say students are happier, too.
Inevitably, AI enthusiasts will criticize school districts that take a cautious approach. Let them. Until results consistently improve, students will be better served by efforts to promote eye-to-eye learning through teacher excellence, high learning standards and classroom accountability.
_____
Michael R. Bloomberg is the founder and majority owner of Bloomberg LP, the parent company of Bloomberg News, and the founder of Bloomberg Philanthropies.
_____
©2025 Bloomberg L.P. Visit bloomberg.com/opinion. Distributed by Tribune Content Agency, LLC.
























































Comments