{"id":4936,"date":"2026-09-29T14:05:35","date_gmt":"2026-09-29T14:05:35","guid":{"rendered":"https:\/\/www.furman.edu\/riley\/?p=4936"},"modified":"2026-09-29T19:08:51","modified_gmt":"2026-09-29T19:08:51","slug":"evidence-matters-ai-in-k-12-education-the-challenging-pace-of-the-frontier","status":"publish","type":"post","link":"https:\/\/www.furman.edu\/riley\/evidence-matters-ai-in-k-12-education-the-challenging-pace-of-the-frontier\/","title":{"rendered":"Evidence Matters | AI in K-12 Education: The Challenging Pace of the Frontier"},"content":{"rendered":"

In Part 1<\/a>, we looked at AI’s promise as a learning partner\u2014and the risk that convenience can replace the productive struggle students need to develop lasting skills.<\/p>\n

But schools face another challenge: we’re adopting AI faster than we can determine what actually works, prepare children for an AI-infused environment, and assess the risk the technology poses.<\/p>\n

We’re Adopting AI Faster Than We Can Study It<\/strong><\/h3>\n

AI is evolving faster than researchers can evaluate it. The Brookings Institution has noted that evidence about which educational AI tools actually improve learning remains limited. Traditional studies can take years, while AI products may change substantially before research is complete.<\/p>\n

A 2026 Brookings report<\/a> argues that education research needs more flexible approaches, including rapid-cycle testing, to keep pace. Schools therefore need to look beyond whether a tool is sophisticated or marketed as personalized learning. The key question is: What does it actually help students learn?<\/p>\n

Research reviewed by the OECD<\/a> similarly finds that technology’s impact depends largely on how it is used, not simply on whether it is present. That matters not only in classrooms, but also in the content children encounter online.<\/p>\n

When AI-Generated Content Becomes the Lesson<\/strong><\/h3>\n

Children are increasingly surrounded by AI-generated content, much of it created without educational oversight. A 2026 investigation by The 74<\/a> found low-quality AI-generated videos aimed at children, including supposedly educational content containing incorrect information and potentially dangerous behavior.<\/p>\n

AI can produce enormous amounts of engaging content quickly, but speed and volume don’t guarantee accuracy or educational value. Students therefore need more than access to information. They need digital literacy\u2014the ability to evaluate who created information, whether it is accurate, and what evidence supports it.<\/p>\n

The Privacy Problem<\/strong><\/h3>\n

Student data raises another concern.<\/p>\n

EdTech platforms can collect information about students’ performance, behavior, and interactions with digital tools. In May 2026, hackers attacked a popular EdTech tool<\/a>, Canvas, and threatened to leak student and teacher data if demands were not met. A June 2026 investigation by The 74<\/a> examined lawsuits involving educational technology companies and questions about parental consent and student data.<\/p>\n

Thus, schools need to evaluate not only whether an EdTech tool works, but also what information it collects and how that information is handled. Schools must simultaneously be asking about the educational value of a particular product and<\/em> considering the implications for student privacy.<\/p>\n

A Bigger Problem Than AI? \u00a0– The Screen-Time Debate<\/strong><\/h3>\n

It’s also important to put today’s AI debate in context.<\/p>\n

A 2026 analysis from the Education Recovery Scorecard<\/a> argues that America’s current “learning recession” began around 2013\u2014years before the pandemic and before generative AI became widespread. AI therefore isn’t the only factor shaping student achievement. Smartphones, social media, and streaming video had already transformed how young people interact with information.<\/p>\n

The challenge is teaching students to navigate that abundance without allowing it to replace concentration, curiosity, and independent thought.<\/p>\n

And, to complicate matters, we know that not all screen time is equal<\/a>. Writing an essay, practicing mathematics, or receiving feedback is different from passively watching videos recommended by an algorithm. Likewise, using AI to critique a student’s work is different from asking it to complete the assignment in the first place.<\/p>\n

The goal, then, isn’t simply to reduce technology use. It’s to distinguish productive technology use from passive or distracting use. But because we are building the plane while flying it\u2014including teachers and researchers themselves just learning and being exposed to the technology\u2014we are currently more reliant on the personal discernment of decision makers than on reliable data and research for deciding if and how to adopt AI in the classroom.<\/p>\n

In Part 3, we’ll look at what a balanced approach to AI and EdTech could look like and which skills students may need most in an AI-enabled world.<\/p>\n

\n

Kelly Gregory is the Riley Institute\u2019s Director for Public Education Partnerships and Projects and previously taught for 11 years in South Carolina public schools. She holds a bachelor\u2019s degree in Psychology and a master\u2019s degree in Special Education. She also holds a National Board certification as an Exceptional Needs Specialist. She can be reached at kelly.gregory0@furman.edu.<\/p>\n","protected":false},"excerpt":{"rendered":"

In Part 1, we looked at AI’s promise as a learning partner\u2014and the risk that convenience can replace the productive struggle students need to develop lasting skills. But schools face […]<\/p>\n","protected":false},"author":300,"featured_media":4953,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[41],"tags":[50,48],"class_list":["post-4936","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-public-education","tag-applied-public-policy","tag-furman-innovation-lab"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/posts\/4936","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/users\/300"}],"replies":[{"embeddable":true,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/comments?post=4936"}],"version-history":[{"count":2,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/posts\/4936\/revisions"}],"predecessor-version":[{"id":4954,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/posts\/4936\/revisions\/4954"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/media\/4953"}],"wp:attachment":[{"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/media?parent=4936"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/categories?post=4936"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.furman.edu\/riley\/wp-json\/wp\/v2\/tags?post=4936"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}