There is a reasonable assumption embedded in how most universities and certification bodies have adopted remote supervision: that deploying it broadly signals institutional rigour. The more exams monitored, the thinking goes, the more defensible the outcomes. It is an intuitive position. It is also, the evidence increasingly suggests, the wrong one.
The problem is not with remote proctoring itself. The technology, applied with precision, does exactly what it promises. The problem is with the scope of its application and the gap between what institutions believe they are achieving with blanket deployment and what is actually happening to the learners on the other end.
The Deployment Logic That Made Sense in 2020
When universities and professional bodies moved en masse to remote delivery during the pandemic, the adoption of remote supervision tools was an emergency response, not a considered policy. Speed was the priority. Many institutions had spent years debating whether online assessment was a good idea. But when the pandemic forced exams online, they had to adopt it almost overnight, before they had the systems, preparation, or student guidance that a careful rollout would normally need.
The problems that surfaced (student complaints, equity concerns, algorithmic false flags) were largely treated as teething issues. What has become clearer in the years since is that many of those problems were not bugs in the implementation. They were signals about deployment scope.
A review by Manit Malhotra and Indu Chhabra, Ensuring academic integrity through automated online exam proctoring: a decade long systematic review, published in Discover Education, synthesised 80 peer reviewed articles from 2014 to 2024 on AI based proctoring systems. The review found that automated proctoring has advanced considerably, but its performance remains uneven across contexts, particularly where hardware variability, environmental conditions and behavioural interpretation intersect with algorithmic flagging. Its conclusion was not that the technology fails, but that its limitations are context dependent. Which is precisely the point institutions have been slow to internalise.
What Blanket Policies Actually Measure
There is a subtler issue at work. When an institution applies remote supervision uniformly across all assessed work, regardless of stakes or format, it is not simply adding a layer of security. It is changing the assessment environment itself, and not neutrally.
Research published in the International Journal of E Learning and Distance Education established a direct relationship between high trait anxiety and exam performance under proctored conditions, finding that the performance gap between high anxiety and low anxiety students widens specifically in online proctored settings. The implication is significant: blanket proctoring policies may be systematically disadvantaging students whose academic ability is not in question, but whose neurological or psychological profile makes surveillance heavy environments harder to perform in.
This is not a peripheral equity concern. A 2024 arXiv preprint on surveillance and disability in online proctored exams, drawing on first hand accounts from students with disability accommodations, documented a cycle in which proctoring induced anxiety increases the very behaviours (fidgeting, eye movement, restroom use) that automated systems flag as suspicious, which in turn deepens anxiety and cognitive load. For this population, the tool designed to ensure fairness is actively undermining it.
The institution, meanwhile, sees a clean audit trail and assumes everything is working.
The Mismatch Between Flexibility and Surveillance
Here is where the argument gets uncomfortable for institutions that have invested heavily in flexible learning frameworks whilst simultaneously rolling out universal proctoring mandates.
Flexible learning, properly understood, is a commitment to removing barriers of time, place, and circumstance from the learning journey. It was designed, in large part, to serve students who cannot conform to traditional delivery: working professionals, carers, students in regional or under resourced locations, those managing chronic health conditions. These are precisely the students for whom a controlled, surveilled exam environment (often requiring a quiet private room, a compliant device, and a stable internet connection) is most burdensome.
A 2024 study by Asil El Galad, Dean Harvey Betts and Nicole Campbell, Flexible learning dimensions in higher education: aligning students’ and educators’ perspectives for more inclusive practices, published in Frontiers in Education, found that flexibility in higher education needs to extend beyond delivery mode into areas such as assessment type, deadlines, grading and course communication. The result is a structural contradiction: students are offered flexibility in how they learn, then assessed under conditions that strip that flexibility away at the moment it matters most.
A 2025 review by Lumbini Barua and Barbara Lockee, Flexible Assessment in Higher Education: A Comprehensive Review of Strategies and Implications, published in TechTrends, reinforces this point. It argues that flexible assessment needs careful design, clear structure and alignment with learning outcomes, rather than being applied as an afterthought or concession. Institutions that treat remote proctoring as the default endpoint of flexible delivery are, in effect, undermining their own stated commitments to inclusive assessment.
What Selective Deployment Actually Looks Like
None of this is an argument against remote supervision. It is an argument for proportionality; matching the level of oversight to the stakes, the context, and the population being assessed.
For high stakes certification exams (professional licensing, regulated qualifications, high tariff summative assessments), the integrity demand is real, and the case for rigorous supervision is robust. Institutions in this space are increasingly turning to dedicated online exam proctoring software as a precision tool for exactly these scenarios. Professional bodies such as the Global Association of Risk Professionals, which in 2025 extended online proctoring options to candidates sitting its sustainability and AI risk certificates, are demonstrating how this can be done well: offering proctored remote delivery as a considered, candidate centred option within a broader assessment architecture, rather than a blanket institutional mandate.
The distinction matters. When an institution or certifying body deploys remote supervision as one component of a risk calibrated assessment strategy (applied to the exams where integrity genuinely cannot be assured by other means), it signals something different to stakeholders than a policy that monitors everything by default. It signals judgement.
The Policy Question Institutions Are Avoiding
The conversation most assessment directors have not yet had with their boards is a straightforward one: which of our assessed tasks actually require this level of oversight, and which do not?
For low stakes formative assessments, coursework submissions, and competency checks that are one component of a broader performance picture, remote surveillance adds procedural complexity without a proportionate integrity dividend. For the exams where results carry genuine consequence (where a fraudulent pass creates real professional or public risk), the calculus is different, and the investment in robust supervision infrastructure is clearly justified.
The Quality Assurance Agency for Higher Education's 2025 Framework for Flexible Assessment explicitly calls on institutions to align assessment flexibility with institutional goals such as widening participation and inclusive practice, and to invest in the infrastructure needed to make that alignment real. Treating remote proctoring as the default answer to assessment integrity, rather than as a precision tool for specific contexts, is a failure of that alignment.
The technology is not the problem. The assumption that more of it, applied more broadly, produces better outcomes is an assumption worth revisiting.