The Journal. - Google Put AI in Schools. Test Scores Plummeted.

Episode Date: October 5, 2026

As Google’s tools become more and more common in American classrooms, students are increasingly relying on AI tools to bypass the hard work of learning. WSJ’s Shalini Ramachandran explores how this trend is impacting students and sparking a backlash from parents and educators. Ryan Knutson hosts. Further Listening: - How YouTube Took Over the American Classroom - The High Schools Going All-In on AI Sign up for WSJ’s free What’s News newsletter. Learn more about your ad choices. Visit megaphone.fm/adchoices

Can you introduce yourself, please? Yes. I'm Shalani Ramachandran, and I'm an investigative reporter at the Wall Street Journal. And as a parent, I feel like you have just been exploding my brain over and over again with your investigations into what's going on with kids in schools this year. Well, thank you. It's been exploding my brain, too. And when I think there's something there, it ends up being far crazier than I imagined when I actually talked to a bunch of parents and teachers and clinicians, but yeah, it's, my brain is also exploding.

So what are you going to explode our brains with today? You know, earlier this year, I did a big investigation about how YouTube had kind of taken over the American classroom. And the process just understood in a much broader way how Google had sort of trained its eyes on children in schools as like a really important critical.

audience. Shalani has been reporting about how deeply embedded Google has become in our nation's schools, from pushing the benefits of Chromebooks to helping make YouTube a cornerstone of lesson plans. And now, as it's really become sort of deeply enmeshed in American education, it has really pushed out its AI tools

to children. So how did this aspect of the story, the fact that Google's Gemini AI, is now in front of every story, student in America just about. How did that get on your radar? When did you notice that? Well, it was kind of as I was talking to parents,

I kept hearing, like, have you noticed that the kids are just plopping all of their answers into the Google search bar and getting AI answers because Google Search has this AI mode, AI overviews and an AI mode chatbot? As Google has become more ubiquitous in the classroom, it's AI, called Gemini, is now everywhere.

Google says this is a big advantage for students, that AI has the potential to enhance education, and that introducing students to AI early is a good way to prepare them for the future. Not everyone is so sure about that. So parents and teachers are really worried that this rollout has helped push many students to become academically and emotionally dependent on AI. And students have talked about it to me, and they've talked about the struggle to come up with the first sentence, struggle to brainstorm and think of an idea because they're so used to sort of farming out their thinking to AI. Welcome to The Journal, our show about money, business, power.

I'm Ryan Knudsen. It's Monday, October 5th. Coming up on the show, how America's students got hooked on Google's AI. This episode is brought to you by IG Private Wealth. When your financial life gets more complex, having a lot. a clear plan matters. IG advisors build personalized, integrated financial plans that connect your investments, tax strategies, retirement income, and estate planning into one coordinated plan. They will help you make confident decisions about your wealth and your future. Get financial advice that puts you at the center. Visit IGprivatewealth.com to find an advisor near you.

In the early 2010s, as schools were looking to bring technology into the classroom, Google had the perfect product to offer them. A lightweight and expensive laptop. Chromebooks were like, this is simple and low cost. And they pushed it out to schools and they said this has the promise of being able to personalize learning. And at this time also, this was kind of like widely considered to be a good thing by most parents and teachers and schools of like, we got to get technology into the classroom. Technology is the future. We need internet access. You know, let's modernize. That's right. And there was a lot to talk at the time about the homework gap and digital divide and like the kids who are lower income not being able to have access to these devices. So it was seen as a good thing. And it was

encouraged by regulators, by politicians, all of that. And as that happened, Chromebooks became even more and meshed into schools because the teachers felt that, well, we got to make sure the kids like know how to use this thing and are familiar with it. These Chromebooks have become a gateway for young students to rely on all kinds of Google apps for their assignments. They use docs to write their essays, slides for presentations. There's Google Classroom, a platform to share assignments and grades. And then, of course, there's the mother of all Google products, Google Search. Now, all of that has been infused with AI. To learn more about Google's rollout of AI, Shalani spoke with more than 60 teachers, school officials, parents, children, and clinicians.

How has Google pitched this to schools? Like, how are they justifying that this is a good thing for students and classrooms? Yeah, well, they say that Gemini is an AI study buddy for students, and they say that they've made a pedagogically tuned model. And what that basically means is like, instead of giving you the answer, it will take you step by step through it. And it'll ask you questions. And to teachers, Google says stuff like that AI can reduce their workload, can deepen metacognition for children, inspire active learning.

They say can level the playing field for struggling in special education students because you can personalize lessons. Google co-sponsors a fellowship program for top school officials where they learn about Google technologies with an emphasis on AI and how to implement it in their districts. And the company also has an influencer network of more than 6,000 teachers who can get certified for attending workshops. Google says these programs are meant to help educators reimagine instruction, so students can experience learning that is, quote, future-ready. So if you're a student, how ubiquitous is Google's Gemini AI? Where are students encountering it? So the one that's the easiest one to access, we've talked about,

it's the Google search bar. So you put something in, you get the AI overviews, and access to this AI-mode chatbot. It'll give you an answer, and then you can keep talking. can do it and it'll keep having conversation. Right. Google AI is sort of injecting itself between the user and search. Those blue links are now pushed way down the page. Right. And most people just engage with the AI answers that Google's providing. That's exactly right. And then separately, there's Gemini that's been integrated in Google classroom. You might see the little diamond

in your docs. The little Gemini logo. Yeah, the little Gemini thing, or in slides. And as of August, Google rolled out a Gemini companion to the partner districts that had enabled Gemini, rolled it out to all K through 12 kids that's just integrated into those classrooms. They put an AI chatbot essentially into the classroom suite. Into the classroom suite, exactly. And then there's other tools too. There's something called photomath. It's a Google-owned app where kids can upload their math assignments, just screenshots,

and it'll just give you the answer. And it'll say it'll step by step, but it gives you the... I mean, you don't even have to beat up a nerd to get him to do your homework anymore, like the olden days. You can just screenshot it. Well, then there's Google Lens, which is integrated into Chromebooks. And that's the same thing. Screenshots, whether it's a translation, whether it's a question for an essay, whether it's math, it'll give you the answer. School districts have the power to deactivate Gemini in some places, like inside Google

classroom. But other features, like the AI overview that shows up at the top of the search page, can't really be shut off unless the school blocks Google altogether. Which means, for students, it's now very difficult to avoid AI, even if they know they're not supposed to use it for certain assignments. One student Shalany spoke to is a 15-year-old named Richard Scott. He had been pretty good at math, you know, previously, and he had done well. But, you know, eighth grade, math was starting to get hard and it was algebra and, you know, there was quadratic equations and things like that. And he realized that he could put those questions into the Google search bar on his school laptop and that essentially acted like an AI chat bot that would spit out the

answers. So he started popping in his algebra questions into the search bar to quickly kind of finish off his daily schoolwork. And did Richard know that he is probably not supposed to be relying on it so much? Yeah, I mean, he knew that he wasn't doing the work, but he couldn't resist the temptation. And then he started leaning on it for other classes too. Like he mentioned how when he needed to label the parts of a volcano or write an essay about Edgar Allen pose the Raven. He wanted a good hook. He asked the Google search bar and it hopefully gave an AI answer. AI made it easy to get through his homework.

Then his grades in math, which was a subject he'd previously. done well at suffered a lot. And he got mostly bees in the quarterly marking periods but got an F on his final. And the impacts of AI went deeper than just the report card. He said, it slowed my thinking down.

He said, when I was in a test and I couldn't use AI, it was hard to think. And there are a number of teachers from various districts in the country who had kind of echoed that. In this past year, when their school districts had sort of open AI policies, I had so many teachers say,

I can't tell you how this was like the first year when I saw kids just struggling to like come up with something, struggling to brainstorm or like struggling to like come up with a first sentence. The number of kids just staring blankly has been different than ever in years past. What do we know at this point about what these AI chat bots and the prevalence of them is doing to student academic performance? So there's a growing body of academic research that shows that student use of student use of. these kind of consumer chatbots doesn't help learning and in fact can actively hurt it. And then there's test scores which are actually kind of showing it in real time. That's after the break. As more and more schools wrestle with the role of AI, we're getting a clearer picture of how it can

impact learning. One data point is a global survey from the Organization for Economic Cooperation and Development, or OECD. It looked at test scores from more than 700. 150,000 teenage students in countries around the world. The study showed that in the U.S., test scores and reading fell 14 points between 2022 and 2025. The survey said there were many factors that could be contributing to this, like the rise of social media and more time spent on screens. But they did have a specific question about student's AI use, and they found that the kids who reported using AI for specific schoolwork tasks, like drafting texts for writing assignments,

had science scores on average of 20 points lower than students who said they never used AI. 20 points lower, wow. And that's equivalent to about a year of schooling. And what does the research show about what is happening to students' brains? Why is it making them perform worse on tests? Yeah, the biggest takeaways are that students seek shortcuts precisely at the moment the work starts to get mentally taxing. At that moment when it's starting to get tough,

is when the students want to take that AI shortcut. Well, I mean, that just seems like common sense. Also, just thinking about, like, when I was a kid, it's like, if the schoolwork got tough, the option was to not finish it and get a bad grade or, like, power through and find a way. But now for students, there's, like, this third option, which is just ask AI.

Exactly. In other words, kids are thinking less. Teachers Shalini spoke with reported this over and over. So teachers have talked about how students take on, like, short stories and films. kind of reek of sameness after their district enabled Gemini. Like they all kind of had the same thing to say about the story, and they actually didn't delve deeper to kind of understand it.

And then another thing is a lot of teachers would tell me about how students would ask Gemini for everything from how to draft a breakup text to asking like how many hours until 2 p.m. at 1130, when there was like a clock in the classroom. And how this teacher in Georgia said they no longer trust what's going on in their own brain. Another concern some researchers have is that while many students start by asking the chat bot about their schoolwork,

that relationship can expand and have a much bigger social and emotional impact on kids. As they start talking to AI and maybe you first go to it for your physics homework and you're like, oh yeah, this thing really helped me. And then you can complain about your friends to this thing and your parents when they're being annoying.

And AI is so sycophantic and like validating. And it sort of makes you think, oh, yeah, I am right. Like, yeah, my parents were really annoying about this thing. And yeah, my friends do suck when they did that thing. There's this, like, frictionless relationship with this AI that I can have. And it's not going to challenge me. It's going to make me feel good. Why should I bother with those friends who are being annoying or being mean to me?

Yeah. And I think as that happens, there's this emotional dependence that can happen as users, whether it's adults or children, want to seek out frictionless relationships, which is something Google researchers have warned about. They've warned that these chatty anthropomorphized AI poses particular risks to children and could cause diminished intellectual engagement and social degradation more generally

because children kind of struggle to distinguish between AI and human interactions. And I actually heard that from teachers too as I talked to them. How has Google responded to all of this? So Google's main response is, that they have created models that encourage critical thinking and has specific protections for users under 18, including you prevent emotional dependence. And they said that their studies show that AI tools paired with teacher-led instruction can help students learn more effectively.

And they said that, you know, teachers around the world use their technology every day because it solves real classroom challenges and helps improve learning outcomes. Google also says it continues to update Gemini to provide more mental health support. and that it works with academic experts, parents, and teens to help ensure its models are safe and appropriate. Meanwhile, parents have begun raising alarms. There are parent groups all over the country that are starting to clamor in school board meetings

with their superintendents and talk about, you know, what are you guys doing about this? Do you have a plan? Is there any evidence that you have that shows these improved learning outcomes? And if there is no evidence, can we wait until there is before we put it in front of the kids? And, I mean, really, there isn't any evidence, is the truth, because this is also new. So all of these tools have been put in front of kids without any really evidence backing that it helps improve learning. Google points to a 2025 trial funded in Sierra Leone, in which it said some students gained over a year's math progress in eight weeks after using Gemini's tutoring mode.

There are some caveats there. They saw that the strongest students benefited the most. And their quote was, this is a common pattern in educational technology. new tools frequently widen achievement gaps rather than close them. Some school districts that last year were encouraging students to learn to use AI are starting to reevaluate. In early September, New York City banned student AI use for one year in classrooms through eighth grade. Los Angeles has also restricted student access to AI. I think what districts really are wrestling with and school officials is they feel like they need to make their kids, their graduates' future ready.

A lot of the messaging from the tech world and from the Googles of the world is that Future Ready means they have to be AI native children. That's kind of the central tension these school districts are wrestling with is that if the world is an AI world, don't we need to like make sure our kids know it and are versatile and understand what it does and what it doesn't do isn't banning it and sort of looking backward and sticking our head in the sand? I've heard some people make the argument that we just need to accept that this technology is here and that it can answer questions for you and that the school needs to change around that

in the same way that math classes changed after the calculator was invented and we could, you know, use that instead of long division. Yeah, I mean, that is definitely an argument that a lot of district officials and Google executives make. And they also talk about Google Maps and GPS. Like, who's memorizing the way to go to places anymore? Who's memorizing phone numbers? So we are offloading our thinking in many ways to many ways to many.

technologies and that's been something that we've been doing over time. The people on the opposing camp see that as sure of course we can use a tool but first we have to teach the kids how to think and come up with an argument, learn how to write, learn how to communicate and like introspect, all the kind of like skills that come from struggling through an essay or struggling through a math problem like the resilience and sort of things that you don't maybe like think about but that is stuff that you got to teach first before you give them the thing that will tell them all the answers. That's all for today. Monday, October 5th. The Journal is a co-production of Spotify and the Wall Street Journal.

Additional reporting in today's episode by Georgia Wells and Molly Reindman. Thanks for listening. See you tomorrow.