Design AI educational systems that fade support and require learner justification to preserve independent problem solving¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The article recommends designing AI-supported instruction that balances external assistance with the establishment of self-control, so that systems are supportive enough while still allowing independent problem solving. Concrete design moves include adaptive hint reduction and pre-response justification prompts: "smart tutors can provide less and less hints as the performance increases, or ask the learners to justify their answers before giving them."

Design Implications¶

Context¶

Requirements¶

• AI tutoring systems capable of adjusting support levels in response to learner performance

Constraints¶

• Systems offering ongoing support can yield immediate performance improvements while reducing learners' chances of self-regulation; efficacy should not be judged on short-term outcomes only

Target Learners¶

• school-age learners
• adolescents
• adult learners

Target Learning Goals¶

• independent problem solving
• self-regulation development

Affordances¶

• Human Ai Co Regulation Framework

Related Strategies¶

• Discourage mental offloading to AI by balancing guidance depth and frequency with application opportunities
• Completion Problems First
• Balance human and automated feedback to optimize writing skill development
• Use learner skill levels to determine when to transition from worked examples to problem solving in tutoring systems

Examples¶

-

Key Sources¶

• S P, Joseph J, Jose M, S M A, N R and Joseph J (2026) Artificial intelligence as a cognitive partner: a developmental framework for human–AI co-regulation in learning. Front. Dev. Psychol. 4:1835258. https://doi.org/10.3389/fdpys.2026.1835258