In early November, Microsoft said it would supply artificial intelligence tools and training to more than 200,000 students and educators in the United Arab Emirates.
Days later, a financial services company in Kazakhstan announced an agreement with OpenAI to provide ChatGPT Edu, a service for schools and universities, for 165,000 educators in Kazakhstan.
Last month, xAI, Elon Musk’s artificial intelligence company, announced an even bigger project with El Salvador: developing an A.I. tutoring system, using the company’s Grok chatbot, for more than a million students in thousands of schools there.
Fueled partly by American tech companies, governments around the globe are racing to deploy generative A.I. systems and training in schools and universities.
Some U.S. tech leaders say A.I. chatbots — which can generate humanlike emails, create class quizzes, analyze data and produce computer code — can be a boon for learning. The tools, they argue, can save teachers time, customize student learning and help prepare young people for an “A.I.-driven” economy.
But the rapid spread of the new A.I. products could also pose risks to young people’s development and well-being, some children’s and health groups warn.
A recent study from Microsoft and Carnegie Mellon University found that popular A.I. chatbots may diminish critical thinking. A.I. bots can produce authoritative-sounding errors and misinformation, and some teachers are grappling with widespread A.I.-assisted student cheating.
Silicon Valley for years has pushed tech tools like laptops and learning apps into classrooms, with promises of improving education access and revolutionizing learning.
Still, a global effort to expand school computer access — a program known as “One Laptop per Child” — did not improve students’ cognitive skills or academic outcomes, according to studies by professors and economists of hundreds of schools in Peru. Now, as some tech boosters make similar education access and fairness arguments for A.I., children’s agencies like UNICEF are urging caution and calling for more guidance for schools.
“With One Laptop per Child, the fallouts included wasted expenditure and poor learning outcomes,” Steven Vosloo, a digital policy specialist at UNICEF, wrote in a recent post. “Unguided use of A.I. systems may actively de-skill students and teachers.”
Education systems across the globe are increasingly working with tech companies on A.I. tools and training programs.
In the United States, where states and school districts typically decide what to teach, some prominent school systems recently introduced popular chatbots for teaching and learning. In Florida alone, Miami-Dade County Public Schools, the nation’s third-largest school system, rolled out Google’s Gemini chatbot for more than 100,000 high school students. And Broward County Public Schools, the nation’s sixth-biggest school district, introduced Microsoft’s Copilot chatbot for thousands of teachers and staff members.
Outside the United States, Microsoft in June announced a partnership with the Ministry of Education in Thailand to provide free online A.I. skills lessons for hundreds of thousands of students. Several months later, Microsoft said it would also provide A.I. training for 150,000 teachers in Thailand. OpenAI has pledged to make ChatGPT available to teachers in government schools across India.
The Baltic nation of Estonia is trying a different approach, with a broad new national A.I. education initiative called “A.I. Leap.”
The program was prompted partly by a recent poll showing that more than 90 percent of the nation’s high schoolers were already using popular chatbots like ChatGPT for schoolwork, leading to worries that some students were beginning to delegate school assignments to A.I.
Estonia then pressed U.S. tech giants to adapt their A.I. to local educational needs and priorities. Researchers at the University of Tartu worked with OpenAI to modify the company’s Estonian-language service for schools so it would respond to students’ queries with questions rather than produce direct answers.
Introduced this school year, the “A.I. Leap” program aims to teach educators and students about the uses, limits, biases and risks of A.I. tools. In its pilot phase, teachers in Estonia received training on OpenAI’s ChatGPT and Google’s Gemini chatbots.
“It’s critical A.I. literacy,” said Ivo Visak, the chief executive of the A.I. Leap Foundation, an Estonian nonprofit that is helping to manage the national education program. “It’s having a very clear understanding that these tools can be useful — but at the same time these tools can do a lot of harm.”
Estonia also recently held a national training day for students in some high schools. Some of those students are now using the bots for tasks like generating questions to help them prepare for school tests, Mr. Visak said.
“If these companies would put their effort not only in pushing A.I. products, but also doing the products together with the educational systems of the world, then some of these products could be really useful,” Mr. Visak added.
This school year, Iceland started its own national A.I. pilot in schools. Now several hundred teachers across the country are experimenting with Google’s Gemini chatbot or Anthropic’s Claude for tasks like lesson planning, as they aim to find helpful uses and to pinpoint drawbacks.
Researchers at the University of Iceland will then study how educators used the chatbots.
Students won’t use the chatbots for now, partly out of concern that relying on classroom bots could diminish important elements of teaching and learning.
“If you are using less of your brain power or critical thinking — or whatever makes us more human — it is definitely not what we want,” said Thordis Sigurdardottir, the director of Iceland’s Directorate of Education and School Services.
Tinna Arnardottir and Frida Gylfadottir, two teachers participating in the pilot at a high school outside Reykjavik, say the A.I. tools have helped them create engaging lessons more quickly.
Ms. Arnardottir, a business and entrepreneurship teacher, recently used Claude to make a career exploration game to help her students figure out whether they were more suited to jobs in sales, marketing or management. Ms. Gylfadottir, who teaches English, said she had uploaded some vocabulary lists and then used the chatbot to help create exercises for her students.
“I have fill-in-the-blank word games, matching word games and speed challenge games,” Ms. Gylfadottir said. “So before they take the exam, I feel like they’re better prepared.”
Ms. Gylfadottir added that she was concerned about chatbots producing misinformation, so she vetted the A.I.-created games and lessons for accuracy before asking her students to try them. Ms. Gylfadottir and Ms. Arnardottir said they also worried that some students might already be growing dependent on — or overly trusting of — A.I. tools outside school.
That has made the Icelandic teachers all the more determined, they said, to help students learn to critically assess and use chatbots.
“They are trusting A.I. blindly,” Ms. Arnardottir said. “They are maybe losing motivation to do the hard work of learning, but we have to teach them how to learn with A.I.”
Teachers currently have few rigorous studies to guide generative A.I. use in schools. Researchers are just beginning to follow the long-term effects of A.I. chatbots on teenagers and schoolchildren.
“Lots of institutions are trying A.I.,” said Drew Bent, the education lead at Anthropic. “We’re at a point now where we need to make sure that these things are backed by outcomes and figure out what’s working and what’s not working.”
Kate Conger contributed reporting.
Natasha Singer is a reporter for The Times who writes about how tech companies, digital devices and apps are reshaping childhood, education and job opportunities.
The post As Schools Embrace A.I. Tools, Skeptics Raise Concerns appeared first on New York Times.
