Confronting Ethical Challenges of AI in Education

Discussing the new challenges recent developments in AI generated output, is highly relevant at the moment. The CENTRAL Network is making use of its agile program planning capabilities and created a high-quality webinar series on the focusing on the impact of AI on the academic world.

The organizers of the webinar series do not set out to find the ultimate answer but help with guidelines and ideas from experts and a diverse audience. With the learnings from keynotes by experts and vivid discussions following a Q&A on the many layers of the topic, the CENTRAL network created another instalment of an inspiring online event.

On the 25^th of May Dr. Sheik Faisal Rashid from DFKI Berlin (German Research Center for Artificial Intelligence), has been talking to us on the topic of "Confronting Ethical Challenges of AI in Education" and how Artificial intelligence has brought transformative opportunities to education and learning. AI can offer personalized learning experiences, automating administrative tasks, and improving student outcomes. Dr. Rashid also argued that it has the potential to polarize us, reinforce human biases, impact vulnerable populations, and violate our privacy. In his talk, he investigated some of the ethical aspects of using AI in education and examine how to better mitigate some of the ethical challenges it poses.

This is a summary of some of the key aspects of the webinar.

Use of AI in Education and Learning

AI is already a part of teaching and learning for students and teachers. This segment provides an overview of possible uses for AI in these areas. The following part will deal with its ethical challenges and how they can be met.

Student Teaching

Using AI to teach students (student-facing)

AI based Applications can facilitate language learning and provide access to language courses and deliver real time automated feedback on pronunciation, comprehension, and fluency. Intelligent tutoring system are following a step-by-step sequence of task and gets individualised instructions of feedback without requiring intervention from the teacher. More advanced systems can automatically adapt to the level of engagement from the participant and tutor on a dialogue-based system that follows conversation in natural languages.

Student Supporting

Using AI to support students learning (student-facing)

AI based learning environments can be used, amongst other things, to help identify individual paths to achieving the desired learning targets. Automatic feedback can be made available for writing assignments. AI supported collaborative learning uses individual work style, past performances, and divides learners into groups with same or a mix of abilities or talents.

Teacher Supporting

Using AI to support the teacher (Teacher-facing)

AI can be used by teachers for evaluation and grading of students. This can be done almost automatically. The AI identifies features such as word, usage, grammar, and sentence structure to grade and provide feedback. Another use of AI for teachers can be moderation of forums. Discussions can be analysed accordingly and provide insights into the activities in the forum and highlight students that may need more help. Chatbots can provide answers to frequently asked questions with simple instructions. This can help teachers cover a portion of their workload and provide them with time and focus on other tasks.

System Supporting

Using AI to support diagnostics or system-wide planning (System facing)

Schools are gathering student date, which can be analysed and used to map out the distribution of resources and who may require additional learning support. Learning analytics, cognitive skills such as vocabulary, listening, spatial reasoning, problem solving, and memory are used to measure and to diagnose learning difficulties. This includes underlying issues, that are hard for teacher to pick up but might be detected early using AI systems. Users can form a competence profile including their education and interests. From this Data in combination with current courses offered, relevant study recommendation can be generated.

Ethical Considerations

There can be four key factors to consider when analysing AI and data and their ethical use. Dr. Feisal presented the main objectives of all of them as follows. All these factors have to be consider when using AI in teaching and research.

Human Agency

Is about the individual’s capability to become a competent member o society. With agency a person can determine their own life choices but also be responsible for their actions. Agency underpins other concepts such as autonomy, self-determination and responsibility.

Fairness

Treat everyone fair in and social organisation. Clear processes need to be determined to create access to opportunities all users. Equity, inclusion, non-discrimination, and fair distribution of rights and responsibilities need be part of this process.

Humanity

Addresses considerations for the people their identity integrity and dignity. Everyone needs to consider the well-being safety and social cohesion, meaningful contact and respect that are necessary for a meaningful human connection. This implies respect for intrinsic value and not see humans as data objects. It is a human-centric approach to AI.

Justified choice

Relates to use of knowledge facts, data to justify necessary or appropriate collective choice in an educational environment. It requires transparency and is based on participatory and collaborative models of decision-making.

Requirements for Ethical use of AI

• Human agency and oversight
  • Fundamental rights, children’s rights, human agency, and human oversight.
• Transparency
  • Traceable, explainable and communication
• Diversity, non-discrimination, and fairnessAccessibility, universal design, the avoidance of unfair bias, and stakeholder participation,
  • Which allows use regardless of age, gender, abilities, or characteristics with a particular focus on students with special needs.
• Societal and environmental wellbeing
  • Sustainability and environmental friendliness, social impact, society, and democracy.
• Privacy and data governance
  • Respect for privacy, quality and integrity of data, and access to data.
• Technical robustness and safety
  • Resilience to attack, security and general safety, accuracy, reliability, and reproducibility.
• Accountability
  • Auditability, minimization and reporting of negative impact, trade-offs, and redress.

We would like to thank our expert Dr. Rashid, for taking the time to share his knowledge with us! Also, we would like to thank the audience and all who participated in the discussion.

This summary conducted by the technical coordinators of the CENTRAL Network. The information this article is based on, has been provided by Dr. Sheikh Feisal Rashid.
If you have any requests or concerns, please be so kind as to contact us at central-network@univie.ac.at

Don’t miss any CENTRAL events, you can sign up for the CENTRAL Newsletter on our website as well.