Introduction:

We may not think about it often, but artificial intelligence (AI) is an integral part of our daily lives. We’ve been using it for years without even realizing it. Whether we’re searching the web, checking our emails, scheduling appointments, asking for directions, or receiving personalized recommendations, AI is always there, quietly working behind the scenes. And during the COVID-19 pandemic, our reliance on AI and internet of things (IoT) has become even more evident. It’s incredible to see how AI has permeated various sectors such as healthcare, education, communication, transportation, and more. Living in the modern world means constantly encountering AI-powered applications.

Digital technologies have brought changes to the nature and scope of education. Versatile and disruptive technological innovations, such as smart devices, the Internet of Things (IoT), artificial intelligence (AI), augmented reality (AR) and virtual reality (VR), blockchain, and software applications have opened up new opportunities for advancing teaching and learning

AI is a branch of computer science that simulates intelligent behavior in computers. It is widely used in science, engineering, technology, and education. AI has various applications in education, such as personalized learning systems, automated assessment systems, and facial recognition systems.

The integration of advanced technologies like artificial intelligence (AI), machine learning (ML), the internet of things (IoT), augmented reality (AR), and virtual reality (VR) has rapidly expanded in the daily routines of children. However, there is still a significant lack of understanding regarding these technologies among the younger population. This discrepancy emphasizes the critical need to educate K–12 students globally, empowering them to participate thoughtfully in the digital evolution of societies. As a result, there has been a heightened emphasis on K–12 education in various research fields, including human-computer interaction, in the past decade.

The assessment presents a thorough evaluation of the latest developments in human-computer interaction, learning sciences, computing education, and child-computer interaction spanning the years 2010 to 2020. Conversely, results highlight a pressing necessity worldwide for interdisciplinary and transdisciplinary research that can consolidate these scattered contributions into a more cohesive area of study and application.

There is a scarcity of literature offering direction on educating K-12 students and teachers about the social, cultural, and ethical implications of AI. While students often engage with ethical concepts related to algorithmic bias, there is a lack of focus on promoting AI and ethics understanding key dynamics. This article consolidates ethical issues surrounding AI in education, discusses teaching methods and available curriculum materials for educating students about AI and ethics, and outlines future research areas and recommendations for educators in K-12 settings. Next, let’s take a moment to consider what lies ahead in terms of research in education and other areas that can help K-12 teachers and students make the most of AI’s advantages while minimizing its disadvantages.

Artificial Intelligence and Applications Defined

The advancement of intelligent machines adapting to human behavior has been hastened by the development of artificial intelligence. The latest progress in computer science has led to the emergence of several definitions and explanations of what constitutes AI systems. One definition highlights AI as the capacity of a digital computer or computer-controlled robot to execute tasks typically done by intelligent beings. This definition underscores the replication of human behavior and awareness. Another definition encompasses the amalgamation of cognitive automation, machine learning, reasoning, hypothesis generation and analysis, natural language processing, and intentional algorithm mutation producing insights and analytics matching or surpassing human capabilities.

The integration of these meanings defines artificial intelligence as the technology responsible for creating systems that can emulate human thinking and interaction to accomplish objectives. AI is predominantly distinguished by diverse applications and advanced computer software, such as recommender systems found in platforms like YouTube and Netflix, personal assistants like Apple’s Siri, facial recognition systems like Facebook’s photo face detection, along with educational apps like Duolingo and Chat GBT. Various sub-fields of AI have been deployed in a multitude of applications to further develop these programs. Notably, evolutionary algorithms and machine learning play a pivotal role in AI’s application in K-12 education.

Algorithms Explained:

The backbone of AI lies in algorithms, which are essential components for its functioning. The development of sophisticated and evolutionary algorithms has been closely intertwined with the history of AI. Algorithms can be defined as a collection of rules or instructions that guide computers in solving problems and attaining specific objectives. Essentially, all computer programs are algorithms, encompassing extensive lines of code that represent mathematical instructions for problem-solving purposes. AI algorithms find application in various domains that are typically associated with human behavior, such as speech and face recognition, visual perception, learning, decision-making, and more. Consequently, algorithms serve as a means to provide instructions for nearly any Artificial Intelligence system or application that can be conceived.

Machine Learning Architecture:

Machine learning, derived from statistical learning methods, utilizes data and algorithms to perform tasks that are typically carried out by humans. Unlike traditional programming, machine learning focuses on enabling computers to act independently, without being provided with specific line-by-line instructions. The core of machine learning lies in exposing learning models to abundant and high-quality data. By analyzing this data, machine-learning algorithms identify patterns and construct models that can predict future values. In simpler terms, machine learning can be seen as a three-step process: data analysis and collection, model building to excel in various tasks, and autonomous execution to achieve desired results without human intervention. The well-known applications of AI, such as recommender systems and facial recognition, owe their existence to the working principles of machine learning.

AI applications benefits Education:

Social networking sites, including Facebook and Google Classroom, serve as a bridge between students and teachers beyond traditional classroom settings. Scholars stress the significance of leveraging social media platforms to extend learning opportunities, keep track of students’ welfare, and nurture student-teacher connections. Numerous researchers have delved into the influence of social media on education, highlighting its ability to enhance student learning, teamwork, and community involvement. Moreover, chatbots powered by AI technology are present on social media platforms to offer natural and conversational interactions, exemplified by the “Pounce” chatbot system at Georgia State University for student assistance.

AI-driven tools like personalized learning systems, automated assessments, facial recognition systems, chatbots, and predictive analytics are becoming more common in K-12 education. They use machine-learning algorithms to assist teachers and students in different ways, such as providing instruction in mixed-ability classrooms and offering detailed feedback on writing assignments. These tools help teachers focus on supporting students during collaborative knowledge-building processes.

ChatGPT

Over the past ten years, artificial intelligence (AI) technology has undergone substantial growth in the field of education. Its effects on educational environments and the formation of educational settings have been evident.

In November 2022, a cutting-edge AI technology was introduced that could potentially impact social and educational sciences. This new AI technology, ChatGPT, has exceeded expectations among both regular individuals and those in the AI industry since its launch. ChatGPT has been shown to possess vast capabilities in abstracting, paraphrasing, translating, editing, generating advanced responses to complex questions, and solving mathematical problems.

Even with its recent debut, ChatGPT has left many in the academic realm feeling both anxious and impressed by its implications for education. This research project aims to offer insights into the prevailing attitudes towards AI technology; such as ChatGPT within the educational sector, examining its implementation and the perceived threats it may present.

Concerns have arisen among academics about students potentially misusing ChatGPT for plagiarism, engaging in fraudulent activities in their assignments, and passing off AI-generated writing as their own. This situation has triggered discussions on the ethical implications of utilizing chatbots for text creation, prompting individuals to contemplate its appropriateness. Legal frameworks have been established to address the authenticity of content generated by ChatGPT.

Chatbots and Social networking Sites: (SNSs)

Social networking sites, including Facebook and Google Classroom, serve as a bridge between students and teachers beyond traditional classroom settings. Scholars stress the significance of leveraging social media platforms to extend learning opportunities, keep track of students’ welfare, and nurture student-teacher connections. Numerous researchers have delved into the influence of social media on education, highlighting its ability to enhance student learning, teamwork, and community involvement. Moreover, chatbots powered by AI technology are present on social media platforms to offer natural and conversational interactions, exemplified by the “Pounce” chatbot system at Georgia State University for student assistance.

It is fair to note that AI has the potential to enhance the educational experiences of students and teachers by helping them tackle instructional challenges. However, it’s important to remember that AI should not replace human interaction. Each student has unique learning styles and requirements. While AI can assist teachers in saving time and enhancing cognitive abilities, it should be viewed as just one tool in their arsenal. To fully benefit from AI and reduce drawbacks, teachers and students must be aware of its limitations, risks, and ethical concerns in education.

What are Personal Learning Systems?

Personalized learning systems, also referred to as adaptive learning platforms or intelligent tutoring systems, are widely recognized and valuable applications of AI in the education field. These systems play a crucial role in supporting both students and teachers by tailoring learning materials to meet individual needs and subjects. Instead of relying on traditional methods like worksheets or textbooks, students can engage with adaptive and interactive multimedia versions of course content, such as chemistry lessons. Research has consistently shown that personalized learning systems lead to higher test scores compared to traditional teacher-led instruction, as evidenced by studies comparing students’ scores on researcher-developed or standardized tests.

Microsoft’s recent report (2018), which surveyed over 2000 students and teachers from Singapore, the U.S., the UK, and Canada, further supports the positive impact of AI on students’ learning progressions. These platforms have the potential to identify gaps in students’ prior knowledge and provide tailored learning tools and materials to support their growth. However, it is important to note that existing platforms primarily focus on modeling learners’ knowledge and cognition, without fully incorporating their social, emotional, and motivational states. With the shift to remote K-12 education during the COVID-19 pandemic, personalized learning systems offer a promising solution for distance learning and have the potential to reshape K-12 instruction in the future.

Systems of Automated Assessment:

Automated assessment systems are rapidly gaining popularity in K-12 education as one of the most promising applications of machine learning. These systems use scoring algorithms to evaluate students’ writing, exams, assignments, and other tasks typically handled by teachers. By providing course support and management tools, assessment algorithms can help reduce teachers’ workload while enhancing their effectiveness and productivity. The beauty of these systems lies in their ability to quickly grade essays, offering students various levels of support.

Major online course providers like Coursera and EdX have integrated automated scoring engines into their platforms to evaluate the writing of numerous students. Additionally, a tool called “Gradescope” has been adopted by more than 500 universities to streamline the scoring and assessment process.

By flagging incorrect answers and marking the correct ones, this tool assists instructors in saving time and effort on manual grading. Unlike numeric assessments that focus on right or wrong answers, automated assessment systems approach essay grading and feedback in a unique manner. Ultimately, these scoring systems have the potential to address the complexities of the teaching environment and support students’ learning journey by providing valuable feedback and guidance for improvement.

Predictive Analytics and Facial Recognition Systems:

Facial recognition software has been implemented to encourage educators to observe and analyze students’ facial expressions to gain a better understanding of their behaviors. This enables teachers to intervene when needed, fostering a more student-centered learning environment and boosting student involvement.

Potential Risks and Ethical concerns of the use of AI Applications in Education:

AI systems bring about ethical challenges and risks that contradict the

marketing narrative of presenting algorithms as unbiased and neutral tools. The truth is, algorithms mirror the values of their creators who hold influential positions. When people develop algorithms, they inadvertently embed society’s historical and systemic biases into the data, resulting in algorithmic bias. Despite the lack of explicit intention, this bias becomes evident in various AI platforms, with gender and racial biases being particularly noticeable.

Ideally, when implementing AI applications in K-12 environments, it is essential to address the diverse types of bias and ethical challenges present. Furthermore, factors like privacy, surveillance, autonomy, bias, and discrimination should be carefully examined. It is worth noting that educators will confront unique ethical dilemmas and hurdles depending on the grade level and age of their students. To combat this, tailored strategies and resources are in place for specific grade levels to detect these differences. An example of this is illustrated in Figure 1 below.

Figure 1.

The use of AI in K-12 education raises a significant ethical concern related to the privacy of students and teachers. Privacy violations occur when individuals disclose excessive personal identifiable information or PII on online platforms. Despite existing legislation and standards for protecting sensitive data, AI-based tech companies often breach data access and security, heightening privacy concerns.

To address these concerns, AI systems request users’ consent to access their personal data. However, many individuals grant consent without fully understanding the extent of the information they are sharing, such as language spoken, racial identity, biographical data, and location. This uninformed sharing undermines human agency and privacy, as AI systems diminish introspection and independent thought.

Scholars have also highlighted the ethical issue of mandating students and parents to use these algorithms in education, even if they explicitly agree to sacrifice privacy. Public schools leave them with no choice but to comply with these systems.

Ethical concerns of AI in K-12 education are also present, particularly regarding surveillance and tracking systems that collect detailed information or (data mining) techniques about students and teachers. AI tracking systems, powered by algorithms and machine-learning models, not only monitor activities but also predict future preferences and actions. These surveillance mechanisms can predict students’ learning performances, strengths, weaknesses, and patterns, potentially invading privacy and limiting participation in learning events.

Autonomy-related issues arise when surveillance systems activate problems, particularly concerning an individual’s ability to act in accordance with their own interests and values. The implementation of predictive systems, driven by algorithms, poses a threat to the autonomy of students and teachers, limiting their capacity to govern their own lives. The use of algorithms to predict individuals’ actions based on their data raises concerns about fairness and personal freedom. Consequently, the risks associated with predictive analysis also include the perpetuation of existing biases, prejudices, and social discrimination.

Bias and discrimination are crucial topics in the conversation about AI ethics in K-12 education. In AI systems, existing biases and power structures are integrated into machine-learning models. Gender bias is clearly evident in language translation tasks, where AI may display biases when translating between gender-specific languages. For instance, Google’s translations of gendered professions underscore societal biases and gender-specific stereotypes present in the data.

Decision-making algorithms in AI for K-12 education, including personalized learning, automated assessment, SNSs, and predictive systems, often exhibit bias. While machine-learning models are designed to improve accuracy and fairness, recent incidents have demonstrated the opposite. In England, the cancellation of A-level and GCSE exams in 2020 resulted in disparities in grade outcomes among students.

Understanding Ethics and Teaching concepts that surround AI in Educational settings.

In order to tackle the ethical issues related to AI use in K-12 education, it is essential to educate students and teachers about these challenges and offer support in dealing with them. To meet this requirement, various research groups and nonprofit organizations have created freely available resources centered on AI and ethics. These resources include teaching materials like lesson plans and interactive activities for students and educators, along with professional development resources such as open virtual learning sessions for teachers. Student awareness early on builds a strong foundation to safe practice and practical understanding behind the dynamics and levels of ethics, privacy concerns, and compliance.

In this blog, we will evaluate three specific resources: the “AI and Ethics” curriculum and “AI and Data Privacy” workshop from MIT Media Lab, and Code.org’s “AI and Oceans” activity. For those interested in exploring additional approaches and resources for AI and ethics in K-12 education, we recommend looking into the AI for the Future Project (AI4Future) by The Chinese University of Hong Kong (CUHK), IBM’s Educator’s AI Classroom Kit, Google’s Teachable Machine, the UK-based nonprofit organization Apps for Good, and Machine Learning for Kids.

Artificial Intelligence and Ethics Curriculum” for High School Students, MIT Media Lab.

The MIT Media Lab team has developed an accessible curriculum on AI and ethics specifically designed for middle school students and teachers. This curriculum aims to provide comprehensive support to teachers in guiding students’ understanding of both the technical aspects of AI systems and the ethical and societal implications associated with them. The curriculum consists of a series of lesson plans and hands-on activities that facilitate the learning process.

In the second phase of the investigation titled “Algorithms as Opinions”, students are encouraged to view algorithms as recipes consisting of a set of instructions that transform an input into an output. Initially, students are tasked with creating an algorithm for making the “best” jelly sandwich and peanut butter. Through this exercise, students delve into the concept of what constitutes the “best” and observe how their personal preferences influence their algorithms. This process allows students to realize that algorithms can serve various purposes and objectives. Subsequently, students engage in the “Ethical Matrix” activity, which builds upon the notion of algorithms as opinions.

In this stage, students review the algorithms they created for the “perfect” peanut butter and jelly sandwich. They discuss what makes a sandwich the “best” considering factors like health, convenience, and taste. Through the use of an ethical matrix, students recognize various stakeholders, such as parents, teachers, or doctors, who have a stake in their sandwich algorithm. This activity demonstrates how the values and perspectives of these stakeholders impact the algorithm. Students use the ethical matrix to identify areas of agreement or disagreement among the stakeholders’ values. The ethical matrix is a useful tool for students to acknowledge the different stakeholders in a system or society and learn how to integrate and apply these values in an ethical manner.

Student then participated in an investigation titled ‘Learning and Algorithmic Bias’, where the biased nature of algorithms is revealed, offering valuable insights. Those involved in this investigation analyze the concept of classification, expanding their understanding in the process. By utilizing Google’s Teachable Machine tool, participants embark on an exploration of supervised machine-learning systems. Their task revolves around training a cat-dog classifier using two distinct datasets. The first dataset emphasizes an over-representation of cats, while the second dataset ensures an equal and diverse representation of both dogs and cats. Through comparing the accuracy of these classifiers, students engage in a thoughtful discussion regarding the fairness of each dataset and its corresponding outcome. This activity effectively prompts students to contemplate the existence of bias in facial recognition algorithms and systems..

As the work study program progresses, students join forces to explore the diverse AI systems that exist on YouTube, such as the recommender algorithm and advertisement matching algorithm. Then, analyze the “YouTube Redesign” task, where they reconstruct YouTube’s recommender system by examining stakeholders, their values, and employing an ethical matrix to evaluate the algorithm’s objectives. Following that, they participate in the “YouTube Socratic Seminar” activity, engaging in a discussion based on a condensed Wall Street Journal article. They deliberate on the impact of stakeholders in suggesting changes to the YouTube Kids app and the appropriateness of features like auto play. Throughout the conversation, they reflect on certain scopes of the project such as: “Which stakeholder possesses the most influential power for initiating change?” and “Have you ever encountered inappropriate content on YouTube? How did you respond?”.
comprehensive review was conducted to analyze the impact and consequences of integrating ChatGPT into educational contexts. The main focus was on investigating the potential benefits and drawbacks that ChatGPT could bring to education. Moreover, the review primarily centered around papers that explored the incorporation of ChatGPT in classrooms, schools, and other educational settings. It also addressed the ethical dilemmas and other challenges that may arise from its implementation

The AI and Ethics curriculum offered by MIT Media Lab is an excellent open-access tool for teachers interact with middle school students and offer an observation to the ethical risks associated with AI. By engaging students in collaborative activities, the curriculum prompts them to grapple with issues like bias, discrimination, surveillance, and autonomy in AI systems.

Artificial Intelligence and Data Privacy” workshop series for K-9 students from the MIT Media Lab.

Students between 7 and 14 can participate in a workshop series by the MIT Media Lab’s Personal Robots Group, which focuses on data privacy education and creating data privacy features. Teachers can access the workshop materials online for free.

The first workshop of the series is titled “Mystery YouTube Viewer: Understanding Data Privacy”. In this workshop, students explore the meanings of privacy and data. They view YouTube’s homepage from the perspective of a mystery user, using video clues to predict the appearance and whereabouts of the characters. This exercise allows students to mimic YouTube algorithms in predicting characters. By engaging with these questions and observations, students reflect on the importance of privacy and boundaries, and how algorithms interpret individuals differently based on their creators.

In the second workshop of the series, participants indulge into the world of advertising with a creative twist. This workshop encourages students to explore the deeper meaning and impact of ads in our lives. Together, they work on designing an advertisement using a common object, aiming to make it as transparent as possible. Throughout the workshop, students also learn about malware, adware, and the various elements of YouTube ads. The participants get to showcase their completed ad posters to their parents and peers.

By engaging with the workshop’s content, students not only learn about the legal frameworks surrounding data privacy but also develop a sense of responsibility towards protecting their own and others’ privacy online. Through hands-on activities and discussions, participants are able to grasp the complexities of privacy issues in the digital realm and brainstorm innovative solutions to address them. Ultimately, the workshop equips students with the knowledge and skills needed to navigate the evolving landscape of social media platforms while upholding privacy rights.

Workshop resources hold great potential in educating both students and teachers on the ethical dilemmas posed by AI in education. One area where these resources can make a significant impact is through the use of social media platforms like YouTube, which are widely utilized as teaching and learning tools in K-12 classrooms and beyond. By incorporating these resources, educators can enhance students’ understanding of data privacy concerns and encourage them to think critically about safeguarding their privacy online. It is crucial for educators to recognize the importance of engaging students in discussions about data ownership, as this can foster a deeper understanding of the underlying reasons behind laws and stimulate debates on their fairness and validity.

Inspecting “AI for Oceans” by Code.org.

When it comes to supporting K-12 educators in addressing the ethical dilemmas surrounding AI, Code.org emerges as a top recommendation. This nonprofit organization, backed by tech giants like Microsoft, Facebook, Amazon, and Google, is focused on fostering students’ participation in computer science. Code.org offers a unique activity called “AI for Oceans,” which allows students in grades 3-12 to train their own machine learning models. By engaging in this activity, students not only learn about AI, algorithms, machine learning, and bias but also develop a deeper understanding of the ethical challenges associated with these technologies. With Code.org’s resources, educators can effectively guide their students through the intricacies of AI ethics and empower them to become responsible users and creators of AI systems.

By actively participating in workshops, students not only gain practical experience in data classification but also develop a deeper understanding of the complexities involved in machine learning. The hands-on nature of the activity allows them to see firsthand how training data influences the outcomes of machine-learning systems and the significance of addressing biases in data collection and analysis. Overall, this handbook serves as a valuable tool for educators looking to empower their students with knowledge and skills in the field of data science and machine learning.

Conclusion:

The goal of this blog is to provide an overview of the potential ethical and societal risks associated with the integration of AI in education. The examples used help emphasize the importance of addressing these risks by offering various instructional strategies and resources for educators who wish to incorporate AI applications into K-12 structure and offer practical readiness and preparedness strategies through active workshops that will educate students about the issues surrounding ethics and data privacy in AI.

Furthermore, upon conducting thorough comparisons between Artificial Intelligence and its integration in K-12 education, it becomes evident that this research blog presents valuable insights into the potential uses of various concepts and ideas. It is clear that there are available tools and credible resources that actively engage and empower teachers and students with logical processes. These processes offer significant advantages in effectively addressing concerns such as personal identifiable information (PII) privacy and bias, while also aiding in their identification and remediation. Moreover, existing research emphasizes the potential of AI to enhance students’ understanding and utilization of corrective AI implementation, highlighting the importance of developing tailored instructional practices for K-12 teachers, specifically focusing on AI, ethics, and privacy.

Moreover, when developing curriculum and designing workshops, it is essential to take into account culturally relevant and responsive pedagogies. This involves centering on students’ funds of knowledge, family backgrounds, and cultural experiences to create instructional materials that effectively tackle issues such as surveillance, privacy, autonomy, and bias. By incorporating these elements, educators can create a more inclusive and engaging learning environment that empowers students to voice their own experiences and perspectives.

Resources:

Timotheou, S., Miliou, O., Dimitriadis, Y. et al. Impacts of digital technologies on education and factors influencing schools’ digital capacity and transformation: A literature review. Educ Inf Technol 28, 6695–6726 (2023). https://doi-org.remote.baruch.cuny.edu/10.1007/s10639-022-11431-8

Akgun, S., Greenhow, C. Artificial intelligence in education: Addressing ethical challenges in K-12 settings. AI Ethics 2, 431–440 (2022). https://doi-org.remote.baruch.cuny.edu/10.1007/s43681-021-00096-7

Huang, Z. IoT-inspired teaching for legal education: AI-based learning based on decision tree algorithm. Soft Comput 28, 1609–1631 (2024). https://doi.org/10.1007/s00500-023-09451-8

‌Kallogiannakis, M., Papadakis, S., Gözüm, A. İ. C., & İpek, Z. H. (2023). Educational Applications of the ChatGPT AI System: A Systematic Review Research. Education Process: International Journal, 12(3), 26–55. https://doi.org/10.22521/edupij.2023.123.2