Trust and ethical considerations in a multi-modal, explainable AI-driven chatbot tutoring system: The case of collaboratively solving Rubik's Cube

Data privacy requires active consent from all stakeholders: The ALLURE chatbot system prioritizes informed consent from children, parents, and teachers before collecting any data, recognizing that AI educational tools must treat consent as foundational rather than assumed, especially when minors are involved.

AI systems need built-in safeguards against harmful content: The platform includes mechanisms to monitor and filter abusive language across text, audio, and visual inputs from both users and the AI itself, and can actively redirect interactions away from dangerous situations—demonstrating that responsible AI requires proactive content moderation rather than reactive responses.

Information isolation prevents unfair advantage between user groups: The system is designed to prevent information leakage about one group of users from influencing another group's experience, addressing fairness concerns by ensuring that collaborative learning doesn't inadvertently create advantages for some students based on data from others.

Explainability builds trust in educational AI: By making the AI's reasoning transparent through multi-modal explanations (likely combining text, visuals, and other formats), the system helps students, parents, and teachers understand how the AI reaches its conclusions, which is essential for building justified trust rather than blind reliance.

Ethical AI in education requires multi-layered design: Rather than treating ethics as an afterthought, the ALLURE system demonstrates that trustworthy AI requires intentional technological design addressing multiple concerns simultaneously—privacy, safety, fairness, and transparency—from the ground up.