# Education Expert Warns AI Is Causing ‘Cognitive Surrender’ Among Students During Senate Hearing

https://www.youtube.com/watch?v=YBajD6T63lE

[00:00] Chairman Tuberville, ranking member
[00:01] Blunt Rochester, and distinguished
[00:03] members of the HELP Committee, thank you
[00:05] for convening this critical hearing.
[00:09] My name is Erin Mote and I'm the CEO of
[00:11] Innovate EDU, a non-profit organization
[00:13] dedicating to ensuring technology serves
[00:17] as a powerful tool for access and
[00:19] opportunity for all students. I come
[00:22] before this committee not only as a
[00:23] charter school founder, but also as a
[00:25] technologist who spent more than two
[00:27] decades working at the intersection of
[00:29] technology policy and systemic change.
[00:32] Yet, perhaps my most important role is a
[00:35] deeply personal one. I'm the mother of
[00:37] two school-aged children.
[00:40] Each day, I see firsthand how these
[00:42] technologies impact real kids at the
[00:45] kitchen table and in classrooms, which
[00:48] fuels my urgency to ensure that our
[00:50] national digital infrastructure is
[00:52] secure, safe, and intentionally built to
[00:56] serve the next generation.
[00:58] To navigate this landscape, we must
[01:01] reject a pervasive false dichotomy.
[01:04] Safety is not the counter polarity to
[01:07] innovation. Safety is not a drag on
[01:09] progress. Rather, safety is a necessary
[01:12] condition for the effective use of AI in
[01:14] education, the very guardrail that makes
[01:17] meaningful, sustainable innovation
[01:19] possible. Without it, we mistake rapid
[01:23] tool adoption for progress.
[01:25] Generative AI is an arrival technology.
[01:29] Like the internet or electricity, it is
[01:31] reshaping how students, educators, and
[01:33] our society live, work, and relate to
[01:36] each other.
[01:37] Data shows that roughly 85% of teachers
[01:40] and 84% of students are already using
[01:43] generative AI in their classrooms. But
[01:46] this transition did not follow a linear
[01:49] path of institutional procurement or
[01:52] risk evaluation. As a result, school
[01:55] districts and families are experiencing
[01:57] severe operational fatigue. More than
[02:00] half of schools have failed to provide
[02:03] any professional development on the safe
[02:05] use of AI.
[02:07] Our schools urgently need federal
[02:09] leadership, structured support, and
[02:11] dedicated funding.
[02:14] A consumer large language model is
[02:16] engineered for platform retention and
[02:18] user satisfaction. To keep users
[02:21] engaged, it defaults to a dangerous form
[02:23] of algorithmics sickofancy. It does not
[02:26] teach, it indulges. It validates
[02:29] incorrect premises and hand delivers
[02:32] answers, often bypassing the productive
[02:34] struggle required for cognitive
[02:36] development. Learning is an active
[02:39] process. True purpose-built edtech
[02:42] scaffolds learning with a student zone
[02:44] of proximal development, much like the
[02:46] AI-driven math accelerators currently
[02:48] being utilized in Alabama to bridge
[02:51] foundational gaps for learners.
[02:53] When we substitute this with
[02:56] consumer-grade systems, we trigger what
[02:58] warring researchers call cognitive
[03:01] surrender. When human psychology meets
[03:03] this incessantly agreeable tech,
[03:06] independent reason is often
[03:08] relinquished. Alarming data shows that
[03:10] when a chatbot intentionally gave
[03:12] incorrect answers, users accepted the
[03:14] error 80% of the time.
[03:17] We are actively placing yet spots in
[03:20] front of neurologically vulnerable
[03:22] youth, conditioning a generation to
[03:24] surrender critical discernment.
[03:27] Moreover, today the average school
[03:29] district accesses nearly 3,000
[03:32] digital tools a year, creating a
[03:34] fragmented, unmanageable landscape for
[03:36] teachers, families, and students. We
[03:39] must transition to evidence-based
[03:40] purchasing anchored in five edtech
[03:43] quality indicators,
[03:45] ensuring tools are safe, interoperable,
[03:48] usable, inclusive, and evidence-based.
[03:51] Meeting a baseline level of evidence,
[03:54] tier four within the federal SF
[03:56] framework, must be a non-negotiable
[03:58] ticket to enter America's classrooms.
[04:01] When less than half of purpose-built
[04:04] edtech tools meet this standard, only 2%
[04:07] of consumer tools do. We must shift our
[04:10] focus to a new ROI, return on
[04:13] instruction.
[04:15] We also must replace the blunt, outdated
[04:17] metric of screen time with screen value.
[04:20] Counting minutes does not tell us if an
[04:22] interaction is meaningful. For instance,
[04:24] screen-based reading is six to eight
[04:26] times less effective for reading
[04:28] comprehension than physical books. Yet
[04:30] for decoding, a skill essential to the
[04:33] science of reading that requires a
[04:35] thousand at bats to achieve fluency,
[04:37] purpose-built AI excels. Furthermore,
[04:41] blanket bans or hourly limits can
[04:43] disproportionately harm the many
[04:45] children with disabilities or who who
[04:47] rely on digital tools for mandated
[04:50] accommodations. We cannot rely on
[04:52] digital avoidance. Congress must ensure
[04:55] that AI is built for our children's
[04:57] safety, learning, and healthy
[04:59] development, while also building tech
[05:01] literacy and digital responsibility in
[05:03] our students. To protect our students
[05:05] without stifling innovation or widening
[05:08] the digital divide, Congress must
[05:10] approach a waterfall approach to safety.
[05:13] This means applying heavy-handed safety
[05:15] controls to high-risk applications at
[05:18] the top, such as public consumer
[05:20] platforms and relationship-stimulating
[05:22] chatbots, while easing the regulatory
[05:24] burden on lower-risk, purpose-driven
[05:26] educational tools. Finally, the federal
[05:29] government must enact immediate
[05:31] structural policy changes. Reconstitute
[05:34] and fully staff the Office of Education
[05:36] Technology. Establish a joint
[05:38] interdisciplinary AI and education
[05:40] federal research agenda and
[05:42] infrastructure through the NSF, IES, and
[05:45] NIH. Protect E-rate and fund
[05:48] non-consensual intimate imagery
[05:50] enforcement agencies like the FTC.
[05:53] We cannot afford a passive wait-and-see
[05:55] approach. We must design digital
[05:57] regulatory ecosystems at the federal and
[06:00] state level that aggressively target
[06:02] genuine harms while keeping human
[06:04] values, ethics, and critical judgment at
[06:07] the center of the American classroom.
[06:09] Thank you, and I look forward to your
[06:11] questions.