Education's real question isn't whether AI was used, but who did the thinking.

Schools often focus on detecting artificial intelligence or deciding whether students have permission to use it. This treats AI as a simple yes-or-no question. Instead, the authors argue that schools must teach students to work alongside smart tools, because future graduates must use technology effectively and wisely in their jobs. To solve this, the researchers created a new design grammardesign grammarA structured set of rules or language for creating and analyzing designs.. This framework helps teachers design and analyze lessons where thinking is purposefully shared between students, peers, teachers, and machines. It looks at how thinking tasks are divided using four actions: constructing, interrogating, assimilating, and outsourcing. It also helps students judge how to split up their mental work responsibly. Rather than asking if someone used AI on a single task, tests should check if that division of work makes sense across time and tasks. Crucially, the plan lists ten reasons why certain abilities must still stay human-only.

Why it matters

Schools currently police AI use with detection and permission rules, but that misses the point. The real challenge is teaching students to deliberately divide thinking between themselves and intelligent tools, because their future jobs will require orchestrating such partnerships. This shifts education from banning or embracing AI to designing learning where the division of cognitive labor is itself the skill.

Questions answered