UoA just adopted the "Two-Lane Approach" to AI in assessments — one lane locks AI out (exams, orals), the other lets you use it (essays, reports, portfolios). Every course now has to tell you which lane each assessment falls into. Here's what it means for your grades.
If you've ever submitted an assignment wondering "am I actually allowed to use ChatGPT for this, or am I about to get flagged for it?", you're not alone, and the University has just changed the rules to try to fix exactly that problem.
Following Senate's endorsement in November 2025, the University of Auckland has formally adopted what it's calling the Two-Lane Approach to assessment and AI. It's not just another policy memo. It's a structural change to how every course has todesign and label its assessments, and by 2027, it'll apply university-wide.
Here's what you need to know.
The problem it's trying to solve
Right now, AI rules in assessment can feel like a lottery. One paper says no AI at all. Another doesn't mention it. A third assumes you're using it but never says so outright. That inconsistency isn't just annoying, it puts students at real risk of accidentally breaching academic integrity rules they were never clearly told about.
The Two-Lane Approach is the University's attempt to end that ambiguity by making the rules explicit, assessment by assessment.
So what are the two lanes?
Lane 1: Controlled assessment.
This is where AI is restricted by default. Think exams, interactive orals, invigilated tests, practical work assessed by observation. The point of Lane 1 is to give a clear, trustworthy picture of what you actually know and can do, without AI in the mix, unless your course specifically says otherwise.
Lane 2: Uncontrolled assessment.
This covers work done outside supervised conditions; essays, portfolios, case studies, reports, presentations, creative projects. Here, AI use is not restricted. You're allowed to use it.
A crucial clarification the University makes explicitly: don't oversimplify this as "Lane 1 = no AI, Lane 2 = anything goes." Lane 2 isn't a free-for-all, you're still expected to use AI critically and responsibly, and academic integrity rules haven'tdisappeared just because the tool is permitted. Using AI to think with is different from using it to think for you, and that distinction still matters in Lane 2.
The actual win here: transparency
The single most important change for you as a student isn't the lane system itself, it's this requirement: every course MUST now tell you, in plain terms, which lane each assessment falls into.
That means no more guessing. No more being penalised for a judgment call you were never given the information to make correctly. If a course doesn't clearly label its assessments by 2027, that's something worth raising with your course coordinator or through student rep channels, it's meant to be baked into how courses are designed, not an optional extra.
What this means for you, practically
-
This is still being shaped. Programme directors and Associate Deans are the ones deciding how the balance and placement of lanes work in your programme. That means there's a real window for student feedback to influence how it lands,through course reps, evaluations, or Student Voice channels, particularly on the open questions above.
The Two-Lane Approach is a genuine attempt to bring clarity to a genuinely messy area. Whether it delivers on that will depend a lot on how consistently it's actually implemented across courses over the next two years, which is exactly the kind of thing worth keeping an eye on and pushing back on if your experience doesn't match the policy on paper.