This study investigates how German secondary school students critically engage with AI-assisted programming tools and perceive ethical responsibilities. It reveals an "AI paradox" where students show strong ethical awareness but integrate AI-generated code without full understanding, with notable gendered differences in usage and trust.

This paper argues that the widespread adoption of frictionless Generative AI risks "cognitive agency surrender" by exploiting human cognitive miserliness and inducing automation bias. It proposes "Scaffolded Cognitive Friction," using Multi-Agent Systems as computational Devil's Advocates to introduce beneficial epistemic tension, alongside multimodal computational phenotyping to measure its effect.

This paper systematically investigates fairness in medical imaging AI benchmarking competitions, revealing significant biases in dataset composition and critical flaws in data accessibility, licensing, and documentation. The findings highlight a disconnect between leaderboard success and clinically meaningful AI, urging for improved transparency and reusability standards.