Most of a classroom doesn't need special scaffolding to use AI safely — they need the Playbook's general rules applied consistently. A minority need something the general rules don't cover on their own. Both are true, and neither cancels the other out.
The advantage: no diagnosis-linked scaffolding hurdle. The core Playbook framework — Traffic Light zones, Mandatory Friction — applies directly, with no accommodation layer needed on top of it.
The risk: invisibility. There's no IEP meeting, no diagnosis conversation, no adult already primed to watch closely — so quiet over-reliance can take hold with zero warning signs. The MIT cognitive-debt findings weren't measured on a diagnosed population; that was a general-classroom result, which is exactly the point.
The guardrail: Mandatory Friction has to be a universal default, not an opt-in accommodation that only activates once someone's raised a flag — because for this group, a flag usually doesn't come until the habit is already formed.
The advantage: systematic, literal, endlessly patient — a strong fit for a mind that organizes the world by pattern rather than social inference.
The risk: emotional attachment. AI's agreeableness can substitute for human connection, especially in high-functioning kids whose fluency makes the risk easy for adults to miss.
The guardrail: watch behavior, not vocabulary — see the Dependency Checklist.
The advantage: executive-function scaffolding — turning overwhelming, multi-step tasks into ordered, time-boxed checklists.
The risk: a novelty-and-avoidance loop, not attachment. A chat window is infinitely responsive, which is exactly what an ADHD brain struggles to disengage from.
The guardrail: time-boxing — a fixed session length before opening the tool, checked against a single stated task.
The advantage: access — text-to-speech and dictation let a student engage with grade-level content their decoding speed would otherwise block.
The risk: skipping the repetitive practice that actually builds decoding and spelling skill.
The guardrail: phonics and spelling instruction stay tool-free; grade-level content access is where AI belongs.
It would be a mistake to treat "general classroom" as a fourth profile sitting alongside the other three, as if it were just one more special case. It isn't — it's the baseline the rest of this Playbook already assumes. The three profiles that follow are specific, evidence-backed departures from that baseline, not parallel categories of equal scope. Leading with the general case first keeps that hierarchy honest.
This list is expanding on purpose, one fully-researched profile at a time rather than a rushed catch-all. Anxiety-related profiles, auditory processing differences, and dyscalculia are next in queue — each will get the same treatment as the profiles above: a distinct advantage, a distinct risk, and a distinct guardrail, not a paragraph bolted onto an existing one.