June 10, 2024
Generative AI in Education: Personalized Learning and Intelligent Tutoring Systems
Generative AI tools have permeated the education sector and are currently the center of a heated debate. On one hand, some educators see generative AI as a disruptive tool that may negatively impact the quality of education by encouraging plagiarism and misinformation. This has reached the point that some learning institutions have banned it altogether.
However, some professors and companies in the education technology (EdTech) sector see things differently. According to analysts from Morgan Stanley Research, the market is focusing primarily on the technology’s negative impact, overlooking its potential to improve learning at all levels [1].
Both of these assumptions are true, but the latter carries more weight. Generative artificial intelligence tools keep getting more accurate, faster, and reliable. Despite their potential for misuse, Gen AI is more of a service than a threat. When utilized ethically, they could help overcome the flaws in the current education system by facilitating a more effective approach to teaching, skill development, and assessment.
This article will explore the role of generative AI in education, evaluating everything from its impact on education to how some educators are using it to improve learning.
Implementing generative AI in education offers vast possibilities for personalized learning and innovative teaching methods. However, its limitations, such as potential for plagiarism and misinformation, must be carefully considered. Balancing these factors is crucial to harness AI’s benefits while maintaining educational integrity. It is also crucial to hire an experienced AI vendor partner to navigate these complexities effectively, said Edwin Lisowski, COO and co-founder at Addepto.
Key Takeaways
- Generative AI can support personalized learning, immediate feedback, lesson planning, content creation, accessibility, and administrative automation.
- Education institutions must address risks involving hallucinations, plagiarism, bias, privacy, cultural misalignment, and excessive dependence on automated tools.
- AI should strengthen teachers’ capabilities and students’ critical thinking rather than replace human instruction, assessment, or judgment.
- Responsible adoption requires clear usage policies, teacher training, student AI literacy, reliable infrastructure, stakeholder engagement, and defined human oversight.
- Institutions should begin with limited, measurable use cases and evaluate learning outcomes before expanding generative AI across courses or departments.
- Scaling requires continuous monitoring of output quality, accessibility, data protection, operating costs, user adoption, and changes in model performance.
How will generative AI affect education?
Evaluating the impact of generative AI in education requires a multifaceted approach. You can’t focus on the benefits alone without considering its potential negative implications, and vice versa.
However, while evaluating the impact of generative AI on education, you should also consider the fact that Generative AI is here to stay. Currently, up to 92% of Fortune 500 companies spanning numerous industries have adopted generative AI [2]
These companies are reaping immense benefits, particularly when it comes to content creation and business intelligence. With the right datasets, Gen AI can identify patterns, trends, and anomalies, enabling data-driven decision-making and a deeper understanding of market dynamics, operations, and customer behavior. [3]
Therefore, it’s quite easy to imagine a world where generative AI could transform the education sector, leading to a more accessible and equitable system.
Here’s a closer look at both the potential positive and negative impacts of general AI in education.
Generative AI: Potential implications
While generative AI has the potential to bring about significant transformations in the education sector, the public should avoid a ‘technological solution’ belief that AI can solve all problems in the education sector. You should also consider other AI implications, like the possibility of misuse in cheating on writing exams and disrupting the social and cognitive dynamics in the classroom.
There’s also the issue of integrating AI ethically and successfully in education. Take countries in the Asia-Pacific region, for instance. Countries like Singapore and China have already established AI in education guidance and policies. However, numerous countries in the region are struggling to meet basic educational needs, let alone adopt Gen AI.
Any country that wants to overcome barriers to generative AI adoption and scaling needs to implement policies and governance frameworks, develop reliable IT infrastructure, provide teacher training, and, most importantly, ensure equitable internet access.
With most Gen AI models being trained on Western datasets, countries with different languages and cultures might struggle to find Gen AI solutions that resonate with local learners.
For instance, Gen AI models trained on primarily Western data can lack cultural and contextual reliance in certain parts of the world. Furthermore, less-desirable Gen AI characteristics like racial and gender biases emanating from the models’ training data could condition learners’ minds, impacting future generations.
You should also consider the risk of learners over-relying on Gen AI for assessments, which could critically undermine the development of critical problem-solving skills. This, combined with Gen AI characteristics like inaccuracies and hallucinations, could further impact the quality of learning.
In order to curb the potential negative implications of generative AI in education, it is vital to prepare both teachers and students for the imminent integration. This means integrating AI-focused curricula for schools and including AI components in teacher education. This way, educators will have the skills necessary to guide ethical and responsible student engagement with generative artificial intelligence tools.
While possible, the full realization of Gen AI’s potential in education requires a nuanced, comprehensive, and rights-based approach. Countries around the world can achieve this by balancing cutting-edge research, policy development, social dialogues, teacher training, and cultural localization efforts.
Generative AI: Potential benefits
While generative AI may have its drawbacks when utilized in learning, you can’t help but look at its overwhelmingly impressive potential benefits. Some of the greatest benefits of Gen AI in education include the following:
Personalized learning
Educators around the world have one common problem – catering to the needs of individual learners. Gen AI, on the other hand, can learn from interactions with learners and adopt educational content to suit individual learning interests, abilities, pace, and styles. It can also provide immediate feedback and support, facilitating more effective learning.
Increased accessibility
Various regions around the world, particularly in the developing world, are grappling with geographical and socioeconomic barriers that hinder education efforts. Gen AI can help overcome some of these barriers by providing high-quality learning resources to learners in underserved regions.
Additionally, Gen AI can be customized to support learning for students with special needs. For instance, Gen AI can assimilate advanced communication and assistive technologies to provide tailored learning experiences.
Automated learning
Teaching is a labor and resource-intensive task. Besides teaching, educators must create syllabi, plan lessons, mark exams, and evaluate each learner’s performance to improve their learning experience.
Gen AI can empower educators by acting as a partner in pedagogical innovation. It can also help save time by streamlining lesson planning, automating administrative tasks, and creating engaging learning activities.
Ultimately, this can go a long way in enriching the educational experience for learners and teachers alike. It can also allow educators to focus more on professional development and teaching.
Generative AI in education: Case studies
Considering how AI has taken hold, not only in industries but also in our daily lives, its inevitability to find its way into the future classroom comes as no surprise. The only question that remains is how educators, learners, and the system as a whole will adapt to the technology.
As players in the education sector find new ways to utilize Gen AI, educators, learners, and industry leaders continue to experiment with the technology to identify safe and valuable applications. Selecting the right generative AI solutions and following scaling best practices can help institutions move from isolated experiments to reliable tools integrated with teaching, assessment, and administrative workflows.
Here are some of the most notable instances where general AI has proven effective in enhancing learning and revolutionizing the education sector as a whole.
Personalized Guidance and Mentorship
Since its inception in 2008, Khan Academy has been at the forefront of improving online learning. The institution recently started utilizing Khanmigo, an internally produced AI teaching assistant. [4]
When utilizing the AI teaching assistant, educators in the institution have noted its remarkable capabilities in providing personalized guidance and mentorship to students – something not all students have access to.
Gen AI can provide individual students with a dedicated counselor, academic coach, career coach, and life coach. What’s even more exciting is these ‘mentors’ can be customized to meet the needs of individual students. For instance, learners can interact with AI-powered versions of literary characters, historical figures, and natural phenomena. For instance, a student struggling in math or physics could interact with an AI-powered erosion of Isaac Newton.
Additionally, the ‘mentors’ are designed to listen actively, engage learners in meaningful conversation, and provide insightful feedback. Therefore, by interacting with these AI-powered characters, learners can gain valuable insights into their personal and academic lives, get guidance on overcoming certain challenges, and develop a growth mindset.
Enhancing Critical Thinking and Analytical Skills
Generative AI can be quite effective in equipping learners with critical thinking and analytical skills. This is clearly evident in use cases that involve interactive activities like debates and collaborative learning exercises. Such activities challenge learners to think beyond the textbook and apply knowledge gained from critical thinking in real-world contexts.
By utilizing generative artificial intelligence, learners can leverage the technology’s capabilities to hone their reading, writing, comprehension, and math skills. Gen AI tools can also provide detailed feedback that helps learners improve. Take this situation, for instance. When a learner submits a writing assignment, generative AI can analyze the text for syntax, grammar, and coherence. It can then provide feedback on areas that may need further improvement.
Updating learning materials
Recent estimates show that about 328.77 million terabytes of data are created each day. [5] This means that the information you have at hand may become outdated almost as soon as you read it. Online resources are being constantly updated with new information and advancements, making it harder for learners and educators to keep up.
Keeping learning resources up to date is a very laborious and time-consuming process. Besides combing through tons of learning materials, you also need to check for any inaccuracies and outdated information before you can embark on the daunting task of updating the materials.
However, with Gen AI, educators and companies in the EdTech sector can streamline the process and make it faster and more effective. What’s even more impressive is that Gen AI applications aren’t limited to textual resources. Gen AI models can also update image resources and will soon be capable of updating and enhancing video resources.
Providing revision and practice questions
Despite the debates around AI’s accuracy and reliability in providing educational resources, you should appreciate the fact that the technology is still young and rap
idly evolving. Additionally, specialized Gen AI models are quite accurate in their domain, thus improving their accuracy and reliability.
Students around the world are already implementing generative AI tools into their learning. With proper guidance, they can do it in a way that does not negatively impact their originality and critical thinking.
Gen AI capabilities like machine learning and natural language understanding enable Gen AI models to analyze education material and generate practice questions and revision resources. They can also be prompted to suggest practice questions based on different levels of understanding, thus improving the revision plan’s effectiveness.
Final thoughts
While there may be ethical concerns about utilizing generic AI tools in education, the technology’s benefits far outweigh its potential drawbacks. However, to achieve ethical and effective adaptation of generative AI in education, players in the education and EdTech sectors must create policies and frameworks to guide effective, ethical adaptation.
It is also important to consider that generative AI technology is rapidly evolving. Some of the latest models are capable of analyzing and generating different data types, including text, images, and videos. Newer models are also more accurate and less prone to hallucinations.
As technology continues to advance, we’re poised to see the development of more reliable systems. Countries around the world will also play a major role in ensuring safe and effective adaptation.
References
[1] Morganstanley.com. Generative AI Is Set to Shake Up Education. URL: http://surl.li/ujfdk. Accessed on June 3, 2024
[2] Explodingtopics.com. Generative AI Stats. URL: https://explodingtopics.com/blog/generative-ai-stats. Accessed on June 3, 2024
[3] Masterofcode.com. Benefits of Generative AI. URL: http://surl.li/ujfeb. Accessed on June 3, 2024
[4] Khanmigo.ai. Khan Academy’s AI-Powered Teaching. URL: https://www.khanmigo.ai. Accessed on June 3, 2024
[5] Explodingtopics.com. Data Generated Per Day. URL: https://explodingtopics.com/blog/data-generated-per-day. Accessed on June 3, 2024
FAQ
Which generative AI use case should an education institution implement first?
The first use case should solve a clearly defined problem, involve limited risk, and produce measurable results. Suitable pilots may include lesson-plan assistance, summarizing approved materials, generating practice questions, supporting administrative communication, or helping students retrieve information from a controlled knowledge base. Institutions should avoid beginning with fully automated grading or other high-impact decisions before governance, evaluation, and human review processes are established. A phased approach makes it easier to test value before investing in broader infrastructure and integrations.
How should schools measure whether a generative AI tool improves learning?
Usage volume or student satisfaction alone does not demonstrate educational value. Institutions should compare outcomes with an established baseline and measure factors such as subject understanding, quality of revisions, retention, task completion, teacher workload, student independence, accessibility, and frequency of incorrect AI-generated guidance. Evaluation should also verify whether the tool supports the intended learning objective rather than merely helping students finish assignments more quickly. NIST recommends testing generative AI systems against defined quality, reliability, and risk criteria throughout their lifecycle.
What should an institutional generative AI policy cover?
The policy should define permitted and prohibited uses, disclosure requirements, responsibilities for checking outputs, acceptable data inputs, intellectual property rules, human review procedures, and consequences for misuse. It should also explain how AI may be used in assignments, research, feedback, grading, and administrative work. UNESCO promotes a human-centred approach that preserves inclusion, equity, critical thinking, and human agency, while current European guidance also highlights the AI Act, GDPR, and responsible use of educational data.
How should student data be protected when using generative AI?
Students and teachers should not enter personal, confidential, medical, behavioural, or assessment information into an external AI tool unless the institution has approved the system and established appropriate data-processing safeguards. Before deployment, the institution should review retention rules, model-training practices, access controls, subcontractors, data location, deletion procedures, and contractual responsibilities. Responsible adoption also requires engagement with affected stakeholders, including teachers and parents, and explicit consideration of user privacy.
What training do teachers need before generative AI is introduced?
Teachers need more than basic prompt-writing skills. Training should cover AI foundations, output verification, privacy, bias, copyright, appropriate classroom use, assessment design, accessibility, and methods for preserving student agency. UNESCO’s framework defines teacher competencies across five areas: a human-centred mindset, AI ethics, AI foundations and applications, AI pedagogy, and AI-supported professional learning.
How can an AI education tool be adapted to local languages and cultural contexts?
Institutions should evaluate the tool using examples written by local educators and learners rather than relying only on generic benchmarks. Testing should examine terminology, cultural references, dialects, curriculum alignment, representation, and performance for different student groups. Local teachers and subject-matter experts should review generated materials and help create approved reference sources. UNESCO emphasizes that educational AI should support inclusion and equity instead of widening existing technological and cultural divides.
Can AI detectors reliably prove that a student used generative AI?
A detector score should not be treated as conclusive proof by itself. Detector performance can vary depending on the model, text length, editing, language, and writing style. NIST continues to evaluate the capabilities and limitations of systems designed to distinguish AI-generated text from human writing, which indicates that this remains an active measurement problem. As a practical inference, institutions should combine any automated signal with assignment history, drafts, oral discussion, source verification, and a fair human review process.
Category:
