What concerns SHOULD we be discussing about using AI in Education?

STEM Teaching Tool #109

-- Topics: Instruction

Equity

TeachClimate

What concerns SHOULD we be discussing about using AI in Education?

Email Feedback

-BACKGROUND

Download as a printable PDF:

Why It Matters to You

Educators should act as critical consumers and users of AI products and services in this moment of hype and promotion. As cultural stewards, they should help all students learn how to critically and ethically deliberate on the use of AI.

Professional Development Providers should support teachers in the informed and responsible adoption of AI systems.

Educational Leaders should resist quick adoption of AI systems until ethical dimensions, risks, and problems have been investigated and resources are in place to support refusal or slow, responsible use.

What Is The Issue?

With the expanding use of different kinds of artificial intelligence (AI) systems in society and in education, it is vital that educators and administrators stop and make time to explore the range of social and ethical concerns posed by these technologies. AI in Education is a quickly evolving landscape with various promising potential benefits, and yet the educational and social implications are too high to not act in a deliberative and ethical manner. Many believe the risks of generative AI outweigh the benefits. Students have the right to question, explore, resist, refuse, or learn how to responsibly and ethically use AI in their lives to support individual, social, and ecological flourishing.

Authors:

Philip Bell | JUNE 2026

Reflection Questions

Are you intentional and cautious in how you explore, resist, or use AI? What ethical ideas and frameworks guide your use?

Do the tools do the disciplinary thinking for students, or are they designed to make them think more? How do you help students to always evaluate online info? Do you only use tech when it advances teaching and learning?

With the opportunities & risks of AI, when should we use it? AI bots ignore evidence. Should we trust them with science?

Things To Consider

There are a range of social and ethical concerns about AI to deliberate on:

Overarching Concerns about the AI Political Economy and Social Power

1. Significant environmental harms include massive water use, supercharging carbon emissions, ecological disruptions, increased air pollution, and negative impacts of quickly expanding data centers

1. Profit seeking and market capture motives of “AI in education” companies can reduce education to a for-profit commercial product

1. Homogenization of culture, thought, and knowledge through linguistic privileging, cultural bias; Global North data bias; narrowing and flattening of curriculum, assessment, and educational goals

1. “Move Fast & Break Things” approach can jeopardize education’s social contract with the public, violate the protective purpose of schooling, disrupt improvement efforts & impair social relationships

Harms that Occur During AI Model Development and Training

1. Intellectual property has been stolen for the training of AI models

1. Unfair labor practices have been documented in “digital sweatshops” as well as significant psychological harm to AI content moderators

1. “Baked-in” biased responses in AI models can produce regressive and marginalizing responses (e.g., that are racist, sexist, xenophobic)

During Use by Students, Teachers, and Educational Leaders

1. Falsehoods are promoted as true (i.e., hallucinations) and misinformation and disinformation are shared through AI systems

1. Adverse health impacts on young people (e.g., impacts on mental health and youth suicide) must be considered and countered

1. Disruption of social learning processes, cognition, and social development (e.g., offloading learning tasks onto AI systems, diminished metacognitive engagement, and diminished creativity)

1. Data privacy & student safety can be compromised; put up a defense

During Use by Teachers and Educational Leaders

1. Deprofessionalization of teaching and ed leadership by offloading tasks to AI systems and diminishing the need for human expertise

1. Encourages automation which is not a suitable replacement for human labor and jobs; education is a fundamentally human endeavor

Attending to Equity

Groups are exploring if AI can support educational equity. However, the risks of using generative AI in education currently outweigh its benefits. We all need to make room for ethical deliberation and informed decision-making about AI in education since it is a “double-edged sword.”

The negative impacts of AI use in society disproportionately impact

non-dominant communities and efforts are needed to promote equity.

Recommended Actions You Can Take

Teach Based on a Social & Ethical Analysis

Teams of educators should engage with the 13 social and ethical concerns, reflect on them, and let it guide their teaching.

Learn how to teach against AI and how to shape a more just sociotechnical future.

AI initiatives and products are outpacing discussion of responsible uses. Use expertise on how people learn and a “go slow and build” approach to reduce harm.

Design AI Uses with Students & Teachers

Engage students with social and ethical concerns to collaboratively make sense of responsible engagement with AI.

Teach: “Science and technology may raise ethical issues for which science, by itself, does not provide answers and solutions.”

Educators should teach about the environmental and social impacts of AI systems using ethical decision-making.

ALSO SEE STEM TEACHING TOOLS

#15 Equity in Science Education

#71 Equity Projects in Science Ed

#97 Climate Justice Learning

Email Feedback

-BACKGROUND

STEM Teaching Tools content copyright 2014-22 UW Institute for Science + Math Education. All rights reserved.

This site is primarily funded by the National Science Foundation (NSF) through Award #1920249 (previously through Awards #1238253 and #1854059). Opinions expressed are not those of any funding agency.

Work is licensed under a Creative Commons Attribution-ShareAlike 4.0 Unported License. Others may adapt with attribution. Funded by the National Science Foundation (NSF). Opinions expressed are not those of any funding agency.