A bill to find out what is not known before AI enters the classroom - workshops directed at the NSF (H.R.10042)
H.R.10042, the Understanding AI in the Classroom Act, would direct the NSF Director to complete workshops within 18 months identifying gaps in data and published findings on how integrating AI into classrooms affects children. It was referred to the House Committee on Science, Space, and Technology on 3 August 2026.
Bill overview (primary data)
- Bill numberH.R. 10042
- TypeHouse Bill
- Congress119th Congress
- Latest actionReferred to the House Committee on Science, Space, and Technology.（2026-08-03）
Key points
- The bill directs the NSF Director to run workshops identifying gaps in data and findings on how classroom AI affects children.
- It neither advances nor restricts adoption: what it asks is to establish what is not known.
- Named subjects include attention span and learning independence, development, and over-reliance or trust miscalibration.
- The text specifies examining differential effects by age, by vulnerability, and by AI characteristics including sycophancy.
- Participation may be sought from researchers, educators, school leaders and parent groups.
- A report is to be submitted to committees in both chambers within six months and published.
1Neither regulating nor promoting, but finding the gaps
Bills on AI in the classroom line up on both sides, advancing adoption and restricting it. This one is neither. What it asks for is finding out what is not known: hold workshops, identify gaps in data and published findings, report and publish. It begins from a view that the material for judgement is insufficient.
2What is to be examined is written out
| Subject named in the text | Content |
|---|---|
| Attention span and learning independence | How a child's capacity to learn for themselves is affected |
| Cognitive, social, emotional, behavioural development | Effects across the facets of development |
| Over-reliance and trust miscalibration | Believing AI too readily, or calibrating trust wrongly |
| Differential effects | By age, by vulnerability, and by characteristics of the AI including sycophancy |
That the text names sycophancy is striking. The property of readily agreeing with the user may carry a different meaning for children - a concern that appears here as statutory language.
3Even who is invited is specified
- 1ResearchersPoint to gaps from the side of knowledge
- 2Educators and school leadersBring what is happening in the field
- 3Parent groupsAdd the view from the household
- 4OthersAnyone the Director determines appropriate
Naming parent groups among participants states a judgement that this subject does not close with researchers and educators. That the report goes to committees in both chambers and is to be published also keeps the result from staying inside Congress.
Of the 120 congressional bills this site holds as of 2026-09-02, five were referred to Education and Workforce and 11 to Science, Space, and Technology. That a subject about classrooms was referred to the science committee rather than the education committee also reflects the character of this bill.
Why it matters
For providers of educational AI, the gap-finding this bill asks for could become the ground for future regulation or procurement requirements. Attention span, learning independence and differential effects from sycophancy also serve as axes for evaluating a product. Because the result is designed to be published, its content may become a premise of market discussion directly.
FAQ
Does the bill regulate AI in classrooms?
What is sycophancy?
Where do the results go?
Sources (primary)
Source: Congress.gov (Library of Congress; U.S. legislative materials, public domain). Links go to the official site.
- Congress.gov (bill page, original)
- H.R. 10042（119th Congress）