NEWS: Ranking Member Blunt Rochester and Chair Tuberville Lead Investigation into AI and K-12 Education

Senators Lisa Blunt Rochester (D-Del.) and Tommy Tuberville (R-Ala.) ranking member and chair of the Senate Committee on Health, Education, Labor, and Pensions' Subcommittee on Education and the American Family,sent a letter to the Government Accountability Office (GAO) requesting an investigation into the use of AI in K-12 education. The letter directs GAO to investigate AI's influence on student achievement, teacher professional development, and use in special education.

In addition to Ranking Member Blunt Rochester and Chair Tuberville, this letter was signed by Senator Tim Kaine (D-Va.).

"We write to request that the Government Accountability Office (GAO) investigate the use of Artificial Intelligence (AI) in K-12 education, with a focus on its impact on student achievement, teacher professional development and preparedness, and its use in special education," the Senators wrote. "While the growth of AI tools in education has promised benefits for both educators and students, early observations raise important concerns as well. As AI adoption in schools grows, it is critical that we better understand how these tools are adopted and their effects to make prudent policy decisions in the future."

"Students are using an array of different AI tools, including generative AI, tutoring chatbots, writing assistants, and other feedback tools that provide explanations, hints, or guidance on assignments," the Senators continued. "A proposed benefit of these AI tools is the personalized and real-time feedback that they can provide for students as they complete scholastic assignments, such as writing or math. Early studies on how these AI tools impact student learning indicate mixed results: while students may experience short-term gains when AI tools help them complete assignments more efficiently, it is unclear whether these gains translate into deeper learning or long-term retention."

"It is clear that AI tools are most effective when they augment, rather than replace, human instruction. Effective personalization in K-12 education depends not only on assessing what a student knows, but also on understanding the relational and affective dimensions of learning," the Senators concluded. "To help us understand the use of Artificial Intelligence (AI) in K-12 education, we ask GAO to conduct investigations assessing the following questions."

The full text of the letter can be found here, and below.

Dear Acting Comptroller General Brown:

We write to request that the Government Accountability Office (GAO) investigate the use of Artificial Intelligence (AI) in K-12 education, with a focus on its impact on student achievement, teacher professional development and preparedness, and its use in special education. While the growth of AI tools in education has promised benefits for both educators and students, early observations raise important concerns as well. As AI adoption in schools grows, it is critical that we better understand how these tools are adopted and their effects to make prudent policy decisions in the future.

Students are using an array of different AI tools, including generative AI, tutoring chatbots, writing assistants, and other feedback tools that provide explanations, hints, or guidance on assignments. A proposed benefit of these AI tools is the personalized and real-time feedback that they can provide for students as they complete scholastic assignments, such as writing or math. Early studies on how these AI tools impact student learning indicate mixed results: while students may experience short-term gains when AI tools help them complete assignments more efficiently, it is unclear whether these gains translate into deeper learning or long-term retention. For instance, take a student using an AI tutoring chatbot to help them with math homework. This tool may help them have an overall more personalized and positive experience while reducing their cognitive load during the learning process. However, the reduced level of reasoning that can occur through the use of these AI tools has been shown to lead to weaker recall of learned information. This raises important questions about the long-term educational effects of using AI tools.

It is clear that AI tools are most effective when they augment, rather than replace, human instruction. Effective personalization in K-12 education depends not only on assessing what a student knows, but also on understanding the relational and affective dimensions of learning. Human teachers can respond to students' emotions, motivation, and social context in ways that current AI tools cannot fully replicate. While AI tools could play a beneficial role in providing detailed individualized feedback and support to students, this reality makes teacher preparation central to whether AI tools can be used effectively to support student learning. Proper supervision of student use of AI tools, and appropriate safeguards, are important.

AI also has the potential to greatly impact educators. Teachers can use AI tools like lesson plan generators, chatbots, and student learning platforms. A survey where participants could select multiple categories, conducted across 303 teachers in the U.S. and India, highlighted that 37% of those teachers described themselves as "beginning to explore" AI, 3% indicated "regularly integrating" AI into their practice, 24% reported creating specific AI activities, 17% described training in AI, and 3% reported non-use of AI. There is early evidence to suggest enthusiasm from teachers about the potential of AI tools to streamline workloads at a time when classroom demands and educator burnout continue to grow. Yet, survey data also reflects fear and apprehension from teachers. A Pew Research Center study found that a quarter of surveyed U.S. teachers feel AI tools are doing more harm than good in K-12. As teachers adopt this technology, and as potential pressure to understand and teach students how to use AI tools grows, it is important that teachers have the supports and training they need to be prepared. Basic AI literacy, continued support and training about how to select AI tools, what purposes AI tools are suited to and not suited to, and how to supervise students using AI are essential to helping teachers leverage these tools in a safe and effective manner that benefits both the educator and their students.

The use of AI tools by and for students with disabilities represents another area that needs continued research and investigation. Such AI tools may include conversational chatbots for students with autism spectrum disorder (ASD), AI-driven annotated text for students with dyslexia, and speech synthesizers for students with speech and language impairments. The adaptive capabilities of these tools may offer new educational opportunities for students with disabilities. Still, with the use of these tools comes notable concern about the privacy and use of student data. Parents and educators are also concerned about whether teachers are adequately prepared to use AI tools and understand the risks associated with them. In particular, there is increasing use of AI, and increasing concern from advocates, in the drafting of Individualized Education Programs (IEPs).

To help us understand the use of Artificial Intelligence (AI) in K-12 education, we ask GAO to conduct investigations assessing the following questions.

1. Exploring the effects of AI on students' critical thinking skills
   • What have researchers and educators observed about the effects of AI on students' critical thinking?
     • Do general use technologies, like large language models, and education-specific technologies that have AI enabled features differ in how they impact student learning?
   • What effects do AI tools have on student academic achievement?
   • How do students use AI in the classroom and on what types of assignments?
   • How frequently and in what ways are students using AI for homework?
   • How are students using AI tools supervised?
   • In what circumstances could or does AI support student development of critical thinking skills and in what circumstances would it risk hindering such development?
2. Extent to which teachers are receiving professional development on the use of AI
   • How are teachers commonly using AI in the classroom, and are teachers receiving professional development designed to better use AI as an educator?
     • Who provides professional learning and training for teachers about the use of AI tools and how AI functions? Do tools have to be approved for use?
   • What guidance or tools have been provided by states and school districts to inform how and if teachers use AI in their classrooms or in specific educational contexts?
     • To what extent does this guidance rely upon evidence-based outcomes?
   • How much are states and school districts investing in AI tools and related professional development?
   • Is AI related instruction being integrated into teacher preparatory programs?
3. The role of AI in special education
   • How are students with disabilities using AI as assistive technology specified in their IEPs?
     • What benefits and challenges have educators observed with its usage in this capacity?
   • How are teachers using AI to draft IEPs? What oversight occurs in such situations?
     • Have teachers using AI to draft IEPs received specific instruction on AI safety, potential legal and privacy concerns, and other related compliance issues?