← Back to Research Radar
Academic Publication Academic Publication

Generative AI without guardrails can harm learning: Evidence from high school mathematics

63
Citations
July 1, 2025
Published Date

Research Abstract & Technology Focus

Generative AI is poised to revolutionize how humans work, and has already demonstrated promise in significantly improving human productivity. A key question is how generative AI affects learning—namely, how humans acquire new skills as they perform tasks. Learning is critical to long-term productivity, especially since generative AI is fallible and users must check its outputs. We study this question via a field experiment where we provide nearly a thousand high school math students with access to generative AI tutors. To understand the differential impact of tool design on learning, we deploy two generative AI tutors: one that mimics a standard ChatGPT interface (“GPT Base”) and one with prompts designed to safeguard learning (“GPT Tutor”). Consistent with prior work, our results show that having GPT-4 access while solving problems significantly improves performance (48% improvement in grades for GPT Base and 127% for GPT Tutor). However, we additionally find that when access is subsequently taken away, students actually perform worse than those who never had access (17% reduction in grades for GPT Base)—i.e., unfettered access to GPT-4 can harm educational outcomes. These negative learning effects are largely mitigated by the safeguards in GPT Tutor. Without guardrails, students attempt to use GPT-4 as a “crutch” during practice problem sessions, and subsequently perform worse on their own. Thus, decision-makers must be cautious about design choices underlying generative AI deployments to preserve skill learning and long-term productivity.
generative without guardrails learning evidence school
Read Full Literature

Correlated Market Trend: Adaptive Learning

Bridging academia to market: The 60-day public search velocity mapping directly to the core technology of this paper. Dashed line represents 7-day moving average.

AI Semantic Synergy Context

Connecting this academic literature to real-world market discussions and products.

crossref.org › academic paper
0%

Generative AI’s Impact on Critical Thinking: Revisiting Bloom’s Taxonomy

The integration of generative artificial intelligence (AI) tools like ChatGPT in education has raised concerns that students may become dependent on AI-generated solutions, potentially stifling the...
crossref.org › academic paper
0%

Beware of metacognitive laziness: Effects of generative artificial intelligence on learning motivation, processes, and performance

Abstract With the continuous development of technological and educational innovation, learners nowadays can obtain a variety of supports from agents such as teachers, peers, edu...
crossref.org › academic paper
0%

The effects of over-reliance on AI dialogue systems on students' cognitive abilities: a systematic review

AbstractThe growing integration of artificial intelligence (AI) dialogue systems within educational and research settings highlights the importance of learning aids. Despite examination of the ethi...
roipad.com › narrative analysis
0%

Gpt

GPT models are advancing with new versions and broader integration, yet generative AI is simultaneously introducing significant risks, including the inability of professionals to detect AI-generate...
crossref.org › academic paper
0%

The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers

No description provided.

Frequently Asked Questions (FAQ)

Curated market intelligence mapped to this research.

What is the core focus of the research titled 'Generative AI without guardrails can harm learning: Evidence from high school mathematics'?

This literature focuses on: Generative AI is poised to revolutionize how humans work, and has already demonstrated promise in significantly improving human productivity. A key question is how generative AI affects learning—namely, how humans acquire new skills as they perfor...

Are there open-source GitHub repositories related to Generative AI without guardrails can harm learning: Evidence from high school mathematics?

Yes, open-source projects like gsd-build/gsd-2 (A powerful meta-prompting, context engineering and spec-driven development system that enables agents to work for long periods of time autonomously...) are actively building upon these concepts.

Which startups are commercializing the technology behind Generative AI without guardrails can harm learning: Evidence from high school mathematics?

Products like Pixel are bringing this to market. Their focus is: Scale performance ads without juggling 7 ad platforms.

What other academic literature is closely related to 'Generative AI without guardrails can harm learning: Evidence from high school mathematics'?

Yes, highly correlated activity was mapped. An entry titled 'Generative AI’s Impact on Critical Thinking: Revisiting Bloom’s Taxonomy' discusses this: The integration of generative artificial intelligence (AI) tools like ChatGPT in education has raised concerns that students may become dependent o...

Are there commercial applications of 'Generative AI without guardrails can harm learning: Evidence from high school mathematics' in market news publications?

Yes, highly correlated activity was mapped. An entry titled 'Gpt' discusses this: GPT models are advancing with new versions and broader integration, yet generative AI is simultaneously introducing significant risks, including th...

Cite this Market Intelligence Report

Reference our AI-mapped synergy between this research and the commercial market to instantly build authority.

Commercial Realization

Startups and Open Source tools heavily associated with the concepts explored in this paper.

  • GitHub
    gsd-build/gsd-2
    A powerful meta-prompting, context engineering and spec-driven deve...
  • GitHub
    magnum6actual/flipoff
    Free split-flap display emulator for any TV. The classic flip-board...
  • Product Hunt
    Pixel
    Scale performance ads without juggling 7 ad platforms
  • Product Hunt
    LLM-Citeops
    Fix 60% scores of AEO and GEO scores without single penny.

Associated Media Narrative