Artificial Intelligence (AI) has a pivotal role in many K-12 educational systems, providing benefits for both students and teachers. We analyze diferent viewpoints on whether AI in education is socially responsible. To best utilize AI's potential and ensure responsible use, it is key for governments to implement policies conducive to AI's adoption within classrooms. In this position paper, we discuss the benefits and limitations AI provides to education as well as the steps needed to responsibly use AI in education in the future.
ers, making it an important tool in modern education.
Thus, AI techniques are being used extensively within
classrooms today [
AI systems can help scale and improve personalized learn- individuals.
ing. For example, the popular studying tool Quizlet uses CV can also greatly benefit teachers. Along with NLP,
AI to help users study more eficiently by creating per- CV can be used to aid in test grading. For multiple choice
sonalized study paths that best address a student’s weak tests, CV can automate the grading process by
detecting where students have bubbled in their answers [
landscape, it is still in its early stages and much more should be done by the governments in their AI policymaking. When creating AI-related policies, they should consult experts in AI given the newness and rapidly evolving nature of the technology. Countries like the US should work towards creating global standards for AI usage, and specifically to establish data-protection policies that would make the school district’s data collection fully public to all parties. By adopting laws that clearly distinguish between which data can be publicly or privately collected, the government would allow for greater trust to occur which is key for further adoption of AI-technologies in the educational sphere.
Additionally, school districts would benefit from policies that provide funding towards specific AI-based technologies, such as automated grading systems or schoolsafety measures. Doing so would benefit teacher resources and allow schools to become more safe for students and teachers. Since many of these AI-based solutions are software-only, they can be implemented and integrated within the pre-existing pipelines within school districts for safety and education. Thus, we believe that implementing such policy to increase funding for AIbased technologies would greatly improve the quality of education across the board.
The implementation of AI in education also carries many
concerns that need to be addressed. In a world with
rapidly advancing technology, data privacy is a problem
that is becoming increasingly harder to solve, especially
in education. Students use various diferent technologies
in an online or hybrid learning setting that leads to an
increase in data collection/usage. While this data is
important for the functionality of AI-based applications, it
can also result in an invasion of the student’s privacy. In
fact, one-third of undergraduate students expressed
concern about their privacy when AI technology was used
in their classrooms [
Another concern of AI is the expanded inequity between wealthy and underprivileged schools. Due to As AI becomes more prevalent within the educational greater resources, funding, and accessibility, wealthier realm, it is important to consider the trustworthiness schools are able to expand AI-based learning faster than of AI usage. The responsible use of AI will allow for underprivileged schools, thus allowing their students maximum benefits for both students and teachers. In the to be more technologically proficient. However, under- future, AI researchers should consider addressing some privileged students may lack technological proficiency, of the concerns that exist within today’s models. A move creating a disadvantage for these students if AI is imple- towards privacy-preserving models and less intrusive AI mented into the school system. would allow for greater trustworthiness and adoption of
AI can also lack self-correction, which is important to such systems within classrooms across the world. At the
improve accuracy. Edtech could aid towards accurate stu- same time, it is important to increase AI literacy among
dent placement in a certain subject. These technologies educational administration to allow for greater usage of
work of the limited data collected while the student uses AI. Thus, socially responsible AI and machine learning
the application, which could lead to improper outcomes. technologies will create a lasting impact on the quality
Generalized algorithms found in AI fail to correctly use and access of education.
small nuances in data, such that they could either inflate These examples of AI are crucial to the understanding
the efect of the nuance or completely overlook it. For ex- and discussion of the efectiveness of the current
impleample, a student could lack vocabulary in a specific topic, mentation of trustworthy AI in education. We find that
and this data could be used to determine the reading level in successful implementations of AI in education, both
of the student [