Hearings to examine the future of K-12 education in the age of artificial intelligence

AI in Education and WorkforceSenate Health, Education, Labor, and Pensions Subcommittee on Education and the American Family · June 16, 2026 · 119th Congress
The Senate HELP Subcommittee on Education and the American Family held this hearing to examine how AI is being adopted in K-12 schools and what federal, state, and local policymakers should do to encourage safe and effective use while protecting students. Begins at ▶ 0:14:54
Transcript
Highlights

Title

K-12 education policy in the age of artificial intelligence

Purpose

The Senate HELP Subcommittee on Education and the American Family held this hearing to examine how AI is being adopted in K-12 schools and what federal, state, and local policymakers should do to encourage safe and effective use while protecting students. Witnesses discussed statewide curriculum and workforce programs in Alabama, safety and privacy frameworks, and Delaware's approach to evaluating AI tools, amid disagreement over what role Congress should play versus states and local districts. Chair Cassidy noted he and the ranking member had asked the Government Accountability Office to launch a study on AI's effects on K-12 education. Begins at▶ 0:14:54

Who spoke

• Chair Bill Cassidy (R-LA)▶ 0:14:54: Opened by describing AI as inevitable in students' lives and stressed that technology cannot replace great teachers▶ 0:15:52; noted he and Sen. Blunt Rochester sent a letter to GAO requesting a study on AI's effects on K-12 education, which GAO has agreed to begin▶ 0:17:20.

• Sen. Lisa Blunt Rochester (D-DE), Ranking Member▶ 0:17:59: Criticized the Department of Education's move to transfer core functions like civil rights protections to other agencies without congressional approval▶ 0:18:13; cited reports that 84% of high school students and 85% of teachers already use generative AI▶ 0:21:24; later described Ohio's pre-ChatGPT use of AI to recommend eliminating 5 million words from state regulatory code▶ 0:57:28, and pressed witnesses on the 2025 PowerSchool breach exposing 62 million students' data▶ 1:04:42.

• Sen. Tommy Tuberville (R-AL), presiding for part of the hearing▶ 0:22:46: Introduced witnesses Jones and Mote▶ 0:22:46▶ 0:29:36; in closing rounds asked about AI skills for graduation▶ 1:30:17, teachers' role▶ 1:32:38, long-term cognitive impact▶ 1:33:29, and whether AI will narrow or widen achievement gaps▶ 1:36:54; adjourned the hearing▶ 1:40:19.

• Josh Jones, CEO of QuantHub▶ 0:23:59: Described Alabama's AI/data literacy partnership, including 100+ paid high school AI/data internships and national rankings behind only California and Virginia in data science enrollment▶ 0:24:52; recommended federal funding reward speed of curriculum updates rather than one-time development▶ 0:26:19, and warned against a patchwork of 50 state AI regulations burdening small businesses▶ 1:11:31.

• Erin Mote, CEO of InnovateEDU / EdSAFE AI Alliance▶ 0:30:37: Argued safety is a precondition for effective AI use, not a drag on innovation▶ 0:31:51; cited a study finding users accepted an AI chatbot's incorrect answers 80% of the time▶ 0:33:42; said the average school district accesses nearly 3,000 digital tools a year and only 2% of consumer AI tools meet a baseline evidence tier▶ 0:34:08▶ 0:34:33; later flagged bias in AI grading and predictive tools as a civil-rights concern raised by groups like the National Parents Union and ACLU▶ 1:08:39.

• Sec. Cindy Marten, Delaware Secretary of Education▶ 0:38:15: Described Delaware's goal of 75% of districts using AI tools under strict data/privacy/human-in-the-loop requirements▶ 0:39:50; detailed Delaware's "AI Assurance Lab" that independently tests vendor AI tools before adoption▶ 0:41:41; said education privacy laws need updating given chatbot risks▶ 1:07:18; stressed teacher professional development as central to safe AI use▶ 1:15:18.

• Sen. Jim Banks (R-IN)▶ 0:44:09: Asked witnesses what specific AI skills schools should teach and how to measure effective teaching of them▶ 0:45:08.

• Sen. Kevin Cramer/Sen. Armstrong (R-ND)▶ 1:00:57: Asked about parental guidance on AI and touted the bipartisan "RACE Act" requiring statewide AI-use standards▶ 1:02:48; asked witnesses about the importance of state-level standards▶ 1:03:07.

• Sen. Tim Kaine (D-VA)▶ 1:16:33: Reflected on influential teachers and argued AI cannot replicate human mentorship▶ 1:17:00; asked witnesses what school boards should weigh before adopting AI tools▶ 1:18:56; noted AI email summaries are frequently inaccurate▶ 1:22:37.

• Sen. Chris Murphy (D-CT)▶ 1:23:30: Argued AI has already done damage to kids' critical thinking, friendship, and moral reasoning▶ 1:23:30; questioned whether AI companies' profit motives could addict even teachers, not just students▶ 1:25:13; proposed an expert-approval process for classroom AI tools▶ 1:27:43.

Key moments

• Chair Cassidy announced that GAO agreed to begin a study on AI's effects in K-12 education, requested jointly with Sen. Blunt Rochester▶ 0:17:20.

• Blunt Rochester condemned the Department of Education's same-day announcement of transferring core functions, including civil rights protections, to other agencies without congressional approval▶ 0:18:13.

• Jones said Alabama ranks third nationally in student data-science enrollment (behind only California and Virginia) and has placed over 100 high schoolers in paid AI/data internships with 50+ employer partners▶ 0:24:52▶ 0:27:37.

• Mote testified that only 2% of consumer-grade AI tools meet a baseline "Tier 4" evidence standard, compared to less than half of purpose-built education tools▶ 0:34:33, and that in one study users accepted a chatbot's deliberately wrong answers 80% of the time▶ 0:33:42.

• Marten described Delaware's AI Assurance Lab, which independently tests vendor AI tools for privacy, bias, and accuracy rather than relying on vendor claims before schools adopt them▶ 0:41:41▶ 0:42:41.

• Blunt Rochester cited that 52% of U.S. school districts experienced a cybersecurity incident in 2025, and the 2024 PowerSchool breach exposed data of 62 million students and 9.5 million teachers — the largest breach of children's educational data in U.S. history▶ 1:04:20▶ 1:04:42.

• Jones warned that navigating 50 different state privacy/AI regulatory regimes creates a significant burden for small AI companies, urging federal action to ease that fragmentation▶ 1:11:31.

• Sen. Murphy pressed on whether AI companies' profit incentives could addict teachers the same way they've been shown to affect children, and proposed requiring AI tools to pass expert approval before classroom use▶ 1:25:13▶ 1:27:43; Mote responded that Delaware, Utah, and Vermont are building state-level "sandboxes" to test tools and generate usable evidence▶ 1:28:22.

• Mote cited Wharton research on "cognitive surrender," noting when AI provides diagnostic support alongside human judgment, student learning gains rose 10 percentage points versus 4.5 points for amplification alone▶ 1:26:39▶ 1:27:08.

• Jones disclosed he had "no opinion" when asked whether federal educational privacy laws need updating, while Mote and Marten both said yes, citing chatbot and consumer-tool risks▶ 1:06:19▶ 1:06:56▶ 1:07:18.

Metadata

CommitteeSenate Health, Education, Labor, and Pensions Subcommittee on Education and the American Family
Chamber / CongressSenate · 119th Congress
DateJune 16, 2026
TypeOpen Hearing
Witnesses
(none listed in event metadata)
Videosenate-isvp
Transcript238 caption blocks · 11,836 words · 1:41:03 runtime
EventCongress.gov 338562