AI Marking Software and Your School's Teaching and Learning Policy

A school's teaching and learning policy is the document that says, in plain terms, how the school expects staff to teach, mark and give feedback. It sits above the individual subject and feedback policies and sets the principles they flow from — that feedback should be timely and purposeful, that assessment should be consistent, that workload should be proportionate. When senior leaders review or rewrite that policy, the way staff actually mark is squarely in scope, and AI marking software is increasingly part of that conversation. The question is not whether to mention it, but how to frame it so the policy stays honest and defensible.

This guide is written for headteachers, deputies and teaching and learning leads shaping the policy. It looks at where AI marking software fits, what it can and cannot claim, and how to write it in so the policy reflects considered practice rather than a bolt-on.

Marking and Feedback Principles Come First

A good teaching and learning policy starts from principles, not tools. It states why the school marks — to move learning forward, to give students something they can act on, to inform planning — and only then describes how. AI marking software belongs under the "how", as one means of delivering the principles the policy already holds, never as a principle in its own right. If your policy says feedback should be specific and timely, the relevant point is that a faster first pass on marking helps staff return work sooner and with clearer, criteria-linked comments — not that the software is a good thing to have.

Framed this way, GradeOrbit reads student work against the school's own marking criteria and produces a proposed mark and feedback for staff to review and confirm. That supports the timeliness and purposefulness the policy asks for, while keeping the professional act of feedback with the teacher. Our guide on AI marking software and school feedback policy reviews covers how this fits the more detailed feedback policy that sits beneath the overarching one.

Consistency of Assessment

Almost every teaching and learning policy commits the school to consistent, accurate assessment — the idea that a piece of work should earn the same mark regardless of which teacher marked it or which evening they did it on. Inconsistency between staff, classes and departments is a familiar weakness, and it undermines both the data a school reports and the fairness students experience.

Because GradeOrbit applies the same criteria to every piece of work, it gives departments a uniform baseline to standardise around. It does not remove the professional moderation a school should do; it strengthens it, by making the starting point consistent rather than dependent on who marked what. A policy can legitimately reference this as a mechanism that supports assessment consistency — provided it also makes clear that moderation and professional judgment remain in place. Our AI marking software for school improvement SLT guide covers how leaders frame this across an improvement plan.

Workload Proportionality

Many teaching and learning policies now include an explicit line on workload — a commitment that marking expectations should be proportionate and sustainable, echoing the DfE's long-standing message that marking is a major driver of teacher workload. If your policy makes that commitment, it should be able to point to how the school delivers on it in practice.

AI marking software is one such practical action. GradeOrbit reduces the mechanical part of marking — the reading, the criteria-checking, the drafting of routine comments — so staff spend more of their time on the professional decisions that need a teacher. Written into the policy, that becomes a concrete way the school honours its own workload principle rather than a vague aspiration. Our guide on AI marking software and the DfE workload reduction toolkit maps this onto the toolkit's themes.

What the Policy Should Not Claim

An honest policy does not overclaim, so it is worth being precise. AI marking is assistive: it produces a first-pass mark and feedback that teachers review, adjust and sign off. The policy should never describe it as "automating" marking, as replacing teacher judgment, or as guaranteeing improved outcomes. The accurate framing is that it reduces the mechanical burden of marking and supports consistency, with professional judgment retained by staff at every stage. Leaders who keep that distinction clear will find the wording far more robust under any external scrutiny, including from Ofsted or from staff and parents who ask how the school uses it.

Data Protection and How It Is Written In

Any tool that processes students' work has to satisfy the data protection questions a policy-conscious leadership team will ask, and your data protection officer should be part of the conversation early. GradeOrbit offers a school tier that operates under a data processing agreement, with institutional controls suited to whole-school rollout, distinct from the privacy-first, never-saved model designed for individual teachers. The right configuration depends on how your school wants to store and return students' work, and it is worth involving your DPO before any pilot so the arrangement is documented properly. A sensible route is to reference the tool as one approved means of marking in line with the policy, then pilot with a department before naming it more widely. Our guide on how schools write an AI marking policy sets out that staged approach.

Bring It Into Your Policy Review

GradeOrbit gives senior leaders a concrete, honest way to deliver the marking, feedback and workload principles a teaching and learning policy already commits to — a faster, more uniform first pass on marking that keeps teachers firmly in charge of the final judgment. It is not a principle to celebrate; it is a means to an end the policy already values.

Explore GradeOrbit to see how AI marking could support your teaching and learning policy, and pilot it with a department this term.