[{"data":1,"prerenderedAt":546},["ShallowReactive",2],{"author-profile-james-gupta":3,"author-blog-james-gupta":26,"author-customers-james-gupta":545},{"id":4,"title":5,"bio":6,"expertise":7,"extension":12,"firstName":13,"image":14,"lastName":17,"links":18,"longBio":20,"meta":21,"name":22,"slug":23,"stem":24,"__hash__":25},"authors\u002Fauthors\u002Fjames-gupta.json","Founder & CEO","Dr James Gupta is the founder and CEO of Synap. He writes about secure online exams, assessment integrity and the technology that lets high-stakes exams run reliably at scale.",[8,9,10,11],"Assessment integrity","Proportionate exam security","Learning science","EdTech & exam infrastructure","json","James",{"url":15,"altText":16},"\u002Fimg\u002Fsynap\u002Fjames-gupta.jpg","Photo of Dr James Gupta, founder and CEO of Synap","Gupta",{"linkedin":19},"https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Fjamesgupta\u002F","Dr James Gupta founded Synap to help organisations deliver secure, scalable online exams without compromising on fairness or candidate trust. With a background in medicine and learning science, he writes and speaks about assessment integrity, proportionate security, and the often-overlooked infrastructure that determines whether high-stakes exams actually hold up under pressure. His work has been featured in FE News, Security Journal UK and other industry publications.",{},null,"james-gupta","authors\u002Fjames-gupta","WYUo3BnEsiX3VgDjcbi48UdgRz0udj3w8EqG3hg45Xk",[27,230,398],{"id":28,"title":29,"author":23,"body":30,"description":208,"extension":209,"headerImage":210,"highlights":22,"isPublished":213,"meta":214,"navigation":213,"path":215,"publishedAt":216,"readingTime":217,"seo":218,"seoDescription":219,"seoTitle":220,"slug":221,"stem":222,"subtitle":223,"topicLabels":22,"topics":224,"__hash__":229},"blog\u002Fblog\u002Fthe-ethics-of-online-exams-proportionate-security-controls.md","The ethics of online exams: designing proportionate security controls",{"type":31,"value":32,"toc":194},"minimark",[33,48,51,56,59,63,66,71,79,83,86,90,93,97,100,138,145,149,157,161,164,184,187],[34,35,36],"p",{},[37,38,39,40,47],"em",{},"This article was originally featured on ",[41,42,46],"a",{"href":43,"rel":44},"https:\u002F\u002Fwww.fenews.co.uk\u002Fexclusive\u002Fthe-ethics-of-online-exams-designing-proportionate-security-controls\u002F",[45],"nofollow","FE News",".",[34,49,50],{},"There is an understandable instinct, when the stakes are high, to secure an online exam by watching as much as possible. If cheating is the risk, surely more monitoring is the answer? But treating surveillance as a dial you simply turn up is both an ethical and a practical mistake. Piling on controls does not automatically make an exam more trustworthy — and it can quietly erode the fairness and confidence that make a result worth anything.",[52,53,55],"h2",{"id":54},"more-surveillance-is-not-more-trust","More surveillance is not more trust",[34,57,58],{},"A heavily monitored exam can feel rigorous while being no more valid than a lightly monitored one. What it reliably does is change the experience for the candidate: it raises anxiety, it treats everyone as a suspect, and it generates a flood of \"anomalies\" that still require human judgement to interpret. The confidence you want in a result comes from sound design and defensible process, not from the sheer volume of data you collected along the way.",[52,60,62],{"id":61},"the-ethical-costs-of-blanket-controls","The ethical costs of blanket controls",[34,64,65],{},"Applying maximum security to every assessment, regardless of what is actually at stake, creates three predictable harms.",[67,68,70],"h3",{"id":69},"it-penalises-the-wrong-people","It penalises the wrong people",[34,72,73,74,78],{},"Intrusive controls fall hardest on candidates who cannot easily conform to them — people with disabilities, atypical movements, caring responsibilities, shared homes or modest equipment. A control that flags them disproportionately is not neutral; it is an ",[41,75,77],{"href":76},"\u002Faccessibility\u002F","accessibility"," and fairness problem dressed up as security.",[67,80,82],{"id":81},"it-invites-legal-and-regulatory-challenge","It invites legal and regulatory challenge",[34,84,85],{},"Collecting biometric data, recording candidates continuously and automating consequential decisions are exactly the practices regulators are scrutinising most closely. An organisation that applies them indiscriminately is taking on legal exposure that is often larger than the integrity risk it set out to manage.",[67,87,89],{"id":88},"it-spends-trust-it-cannot-easily-rebuild","It spends trust it cannot easily rebuild",[34,91,92],{},"Candidate trust, once lost, is expensive to recover. An exam process that feels invasive damages the relationship between an institution and the people it assesses — and that reputational cost outlasts any single exam sitting.",[52,94,96],{"id":95},"proportionate-controls-that-scale-to-the-stakes","Proportionate controls that scale to the stakes",[34,98,99],{},"The ethical and effective approach is proportionality: match the security to the genuine risk of the specific assessment. That starts with a simple question asked honestly for each exam — what is actually at stake here, and what threats realistically apply?",[101,102,103,117,127],"ul",{},[104,105,106,107,111,112,116],"li",{},"A ",[108,109,110],"strong",{},"low-stakes, formative"," check needs little more than good ",[41,113,115],{"href":114},"\u002Fonline-exam-platform\u002Fauthoring\u002F","assessment design"," and sensible settings.",[104,118,106,119,122,123,126],{},[108,120,121],{},"moderate-stakes"," exam might add randomised ",[41,124,125],{"href":114},"question banks",", timing controls and access restrictions.",[104,128,106,129,132,133,137],{},[108,130,131],{},"high-stakes"," licensing or certification exam may justify ",[41,134,136],{"href":135},"\u002Fonline-exam-platform\u002Fproctoring\u002F","full proctoring and lockdown"," — applied transparently, with the minimum data necessary.",[34,139,140,141,144],{},"The goal is the ",[37,142,143],{},"least intrusive control that addresses the real threat",", not the most intrusive control available.",[52,146,148],{"id":147},"building-integrity-into-the-design","Building integrity into the design",[34,150,151,152,156],{},"Proportionality works best when integrity is designed into the assessment rather than policed at the surface. Well-constructed items, large and varied question pools, meaningful time limits and thoughtful exam structure reduce the opportunity to cheat before any monitoring is involved. Get the ",[41,153,155],{"href":154},"\u002Fsolutions\u002Fonline-exam-platform\u002F","design"," right and you need less surveillance to achieve the same integrity — which is better for everyone.",[52,158,160],{"id":159},"privacy-by-design-transparency-and-governance","Privacy by design, transparency and governance",[34,162,163],{},"Finally, how controls are implemented matters as much as which controls you choose:",[101,165,166,172,178],{},[104,167,168,171],{},[108,169,170],{},"Privacy by design:"," collect only what the stakes justify, and be deliberate about what you decline to collect.",[104,173,174,177],{},[108,175,176],{},"Transparency:"," tell candidates what is monitored and why. Consent given with understanding costs far less trust than monitoring that feels hidden.",[104,179,180,183],{},[108,181,182],{},"Governance:"," keep a human in the loop for decisions that affect a result, with a clear appeals route and an audit trail behind every judgement.",[34,185,186],{},"Credible digital assessment does not come from watching everything. It comes from watching the right things, proportionately, with candidates' trust and dignity treated as part of the design rather than a casualty of it. That is not a softer position on integrity — it is a more honest one.",[34,188,189],{},[37,190,39,191,47],{},[41,192,46],{"href":43,"rel":193},[45],{"title":195,"searchDepth":196,"depth":196,"links":197},"",2,[198,199,205,206,207],{"id":54,"depth":196,"text":55},{"id":61,"depth":196,"text":62,"children":200},[201,203,204],{"id":69,"depth":202,"text":70},3,{"id":81,"depth":202,"text":82},{"id":88,"depth":202,"text":89},{"id":95,"depth":196,"text":96},{"id":147,"depth":196,"text":148},{"id":159,"depth":196,"text":160},"This article was originally featured on FE News.","md",{"url":211,"altText":212},"\u002Fimg\u002Fblog\u002Fproportionate-security.svg","Illustration of exam security controls scaling in proportion to the stakes of an assessment",true,{},"\u002Fblog\u002Fthe-ethics-of-online-exams-proportionate-security-controls","2026-06-26","6 min read",{"title":29,"description":208},"Piling on surveillance does not make an online exam more trustworthy. How proportionate, privacy-by-design security controls that scale to the stakes protect integrity, fairness and trust.","The ethics of online exams: proportionate security controls","the-ethics-of-online-exams-proportionate-security-controls","blog\u002Fthe-ethics-of-online-exams-proportionate-security-controls","Why matching exam security to actual risk protects both integrity and candidate trust better than blanket surveillance.",[225,226,227,77,228],"exam-integrity","proctoring","security","exam-design","W4Ozqa3Fr3_FvnqOiCJGS02qumqdMs6NRg0UXxmutu4",{"id":231,"title":232,"author":23,"body":233,"description":208,"extension":209,"headerImage":381,"highlights":22,"isPublished":213,"meta":384,"navigation":213,"path":385,"publishedAt":386,"readingTime":217,"seo":387,"seoDescription":388,"seoTitle":232,"slug":389,"stem":390,"subtitle":391,"topicLabels":22,"topics":392,"__hash__":397},"blog\u002Fblog\u002Fthe-infrastructure-blind-spot-in-online-exams.md","The infrastructure blind spot in online exams",{"type":31,"value":234,"toc":371},[235,243,250,254,257,261,269,273,276,280,288,292,295,299,302,334,342,346,361,364],[34,236,237],{},[37,238,39,239,47],{},[41,240,46],{"href":241,"rel":242},"https:\u002F\u002Fwww.fenews.co.uk\u002Fexclusive\u002Fthe-infrastructure-blind-spot-in-online-exams\u002F",[45],[34,244,245,246,249],{},"When organisations set out to choose an ",[41,247,248],{"href":154},"online exam platform",", the conversation almost always starts in the same place: question types, proctoring, marking and reporting. Those things matter, and they demo well. But in my experience they are rarely the reason an exam goes wrong on the day. The thing that decides whether a high-stakes exam succeeds or fails is usually invisible on a feature-comparison sheet — the infrastructure underneath it.",[52,251,253],{"id":252},"why-online-exams-are-an-unusual-workload","Why online exams are an unusual workload",[34,255,256],{},"It is tempting to assume that if a platform can serve a website, it can run an exam. It can't, or at least not reliably. Online assessment places demands on infrastructure that most software never has to handle.",[67,258,260],{"id":259},"the-traffic-arrives-all-at-once","The traffic arrives all at once",[34,262,263,264,268],{},"A cohort does not trickle in. Thousands of candidates click \"start\" within the same sixty seconds, then stay connected — saving answers, loading media, syncing state — for hours. That is a very different shape of load from a normal web app, and a platform that has not been engineered for ",[41,265,267],{"href":266},"\u002Fuse-case\u002Flarge-scale-high-concurrency\u002F","high concurrency"," will show it exactly when it can least afford to.",[67,270,272],{"id":271},"failure-mid-exam-is-not-recoverable","Failure mid-exam is not recoverable",[34,274,275],{},"For most software, a brief outage is an inconvenience. In an exam, it is a candidate who loses their work, a cohort whose results are challenged, and an awarding body whose credibility takes the hit. The stakes change what \"good enough\" means. Resilience — graceful handling of a dropped connection, a browser crash, a flaky network — has to be designed in, not bolted on afterwards.",[67,277,279],{"id":278},"every-candidates-environment-is-different","Every candidate's environment is different",[34,281,282,283,287],{},"You do not control the devices or networks your candidates use. Some sit in a managed test centre; others are at home on a laptop and a mobile hotspot, sometimes ",[41,284,286],{"href":285},"\u002Fuse-case\u002Fchallenging-environments\u002F","in genuinely challenging conditions",". The platform has to hold up across all of it, which is a much harder engineering problem than serving a predictable, well-provisioned audience.",[52,289,291],{"id":290},"why-infrastructure-gets-overlooked","Why infrastructure gets overlooked",[34,293,294],{},"The honest answer is that features are easy to evaluate and resilience is not. You can see a proctoring dashboard in a demo. You cannot see how a platform behaves when ten thousand people start at once, or what happens when a data centre has a bad day — until it happens to you. So the visible layer gets the attention, and the layer that actually determines the outcome gets a line in the contract, if that.",[52,296,298],{"id":297},"questions-worth-asking-a-vendor","Questions worth asking a vendor",[34,300,301],{},"If infrastructure is the blind spot, the fix is to shine a light on it during procurement. A few questions tend to be revealing:",[101,303,304,310,316,322,328],{},[104,305,306,309],{},[108,307,308],{},"Concurrency:"," how many candidates can genuinely start and run at the same time, and how is that number established?",[104,311,312,315],{},[108,313,314],{},"Scaling:"," does capacity scale automatically for a big exam day, or does someone have to provision it in advance?",[104,317,318,321],{},[108,319,320],{},"Resilience:"," what happens to a candidate's attempt if their connection drops or their browser crashes mid-exam?",[104,323,324,327],{},[108,325,326],{},"Durability:"," how often is work saved, how are backups handled, and what are the recovery objectives if the worst happens?",[104,329,330,333],{},[108,331,332],{},"Track record:"," is there a public status page, and what does the incident history actually show?",[34,335,336,337,341],{},"Vague answers to these questions are themselves an answer. This is also where a serious ",[41,338,340],{"href":339},"\u002Ftechnical\u002F","security and data-protection posture"," tends to correlate with a serious infrastructure posture — the organisations that invest in one usually invest in the other.",[52,343,345],{"id":344},"infrastructure-design-and-verification-are-layers-of-trust","Infrastructure, design and verification are layers of trust",[34,347,348,349,352,353,356,357,360],{},"None of this is an argument that infrastructure is the ",[37,350,351],{},"only"," thing that matters. It is an argument that it is a necessary foundation the others sit on. Dependable infrastructure, ",[41,354,355],{"href":114},"sound assessment design"," and ",[41,358,359],{"href":135},"identity verification"," are complementary layers of trust. A beautifully designed, well-proctored exam still fails if the platform falls over when the cohort logs in. And robust infrastructure does not, on its own, make an exam fair or valid.",[34,362,363],{},"The takeaway is not to stop caring about the visible features. It is to give the invisible layer the same scrutiny — because when a high-stakes exam is running and something goes wrong, it is almost never the question types that let you down.",[34,365,366],{},[37,367,39,368,47],{},[41,369,46],{"href":241,"rel":370},[45],{"title":195,"searchDepth":196,"depth":196,"links":372},[373,378,379,380],{"id":252,"depth":196,"text":253,"children":374},[375,376,377],{"id":259,"depth":202,"text":260},{"id":271,"depth":202,"text":272},{"id":278,"depth":202,"text":279},{"id":290,"depth":196,"text":291},{"id":297,"depth":196,"text":298},{"id":344,"depth":196,"text":345},{"url":382,"altText":383},"\u002Fimg\u002Fblog\u002Finfrastructure-blind-spot.svg","Illustration of an exam platform's layers resting on a highlighted infrastructure foundation, with a concurrency spike",{},"\u002Fblog\u002Fthe-infrastructure-blind-spot-in-online-exams","2026-05-21",{"title":232,"description":208},"Organisations choosing an online exam platform focus on question design and proctoring, but reliability at scale is decided by the infrastructure underneath. Here's what to look for.","the-infrastructure-blind-spot-in-online-exams","blog\u002Fthe-infrastructure-blind-spot-in-online-exams","The technical layer that keeps high-stakes exams running is too often ignored when organisations choose a platform.",[393,394,395,396],"online-exams","exam-delivery","scalability","high-stakes-exams","PL00iQhBd3YwUQQQzstRw56lmuqUrgRD8S51ni74Vjo",{"id":399,"title":400,"author":23,"body":401,"description":529,"extension":209,"headerImage":530,"highlights":22,"isPublished":213,"meta":533,"navigation":213,"path":534,"publishedAt":535,"readingTime":217,"seo":536,"seoDescription":537,"seoTitle":538,"slug":539,"stem":540,"subtitle":541,"topicLabels":22,"topics":542,"__hash__":544},"blog\u002Fblog\u002Fbalancing-integrity-trust-and-surveillance-in-exams.md","Balancing integrity, trust and surveillance in exams",{"type":31,"value":402,"toc":518},[403,412,420,424,427,431,434,438,441,445,451,455,458,462,465,475,479,482,504,508,511],[34,404,405],{},[37,406,39,407,47],{},[41,408,411],{"href":409,"rel":410},"https:\u002F\u002Fsecurityjournaluk.com\u002Fbalancing-integrity-trust-surveillance-exams\u002F",[45],"Security Journal UK",[34,413,414,415,419],{},"There is an assumption baked into a lot of online exam security: that if some monitoring is good, more must be better. Turn on the webcam, record the screen, track the eyes, lock down the machine, flag every anomaly. It feels rigorous. But more surveillance does not automatically produce more ",[41,416,418],{"href":417},"\u002Fonline-exam-platform\u002Fsecurity\u002F","integrity"," — and past a certain point it starts to cost you the very trust you were trying to protect.",[52,421,423],{"id":422},"monitoring-is-not-the-same-as-integrity","Monitoring is not the same as integrity",[34,425,426],{},"Integrity is the outcome you actually want: results that reflect what a candidate genuinely knows and can do. Surveillance is one of several means to that end, and a fairly blunt one. A wall of video and behavioural flags generates a lot of data, but data is not evidence, and evidence is not the same as a fair, defensible decision. Piling on monitoring can create an impression of rigour while doing very little to change whether the result is trustworthy.",[52,428,430],{"id":429},"the-costs-of-over-monitoring","The costs of over-monitoring",[34,432,433],{},"Excessive surveillance is not free, even when the software is. It carries real costs that rarely appear in the business case.",[67,435,437],{"id":436},"candidate-trust","Candidate trust",[34,439,440],{},"Candidates who feel watched and distrusted do not perform as themselves. Intrusive monitoring raises anxiety, and anxious candidates make more of the innocent movements that then get flagged as suspicious — a feedback loop that undermines both the experience and the data.",[67,442,444],{"id":443},"fairness-and-accessibility","Fairness and accessibility",[34,446,447,448,450],{},"Blanket surveillance tends to fall hardest on the people least able to conform to it: candidates with disabilities, atypical movement patterns, poor lighting, shared living spaces or older devices. What looks like a neutral control is often a source of bias, creating ",[41,449,77],{"href":76}," problems that quietly skew who passes and who gets flagged.",[67,452,454],{"id":453},"legal-and-regulatory-exposure","Legal and regulatory exposure",[34,456,457],{},"Regulators are paying closer attention to intrusive proctoring — biometric processing, always-on recording, automated decisions about people's futures. Collecting the maximum amount of personal data on every candidate, regardless of the stakes, is exactly the posture that attracts scrutiny and complaints. The compliance risk of over-monitoring is now often larger than the integrity risk it was meant to address.",[52,459,461],{"id":460},"risk-based-proportionate-security","Risk-based, proportionate security",[34,463,464],{},"The alternative is not to abandon security. It is to match the controls to the actual risk of each assessment. A low-stakes practice test, a formative quiz and a licensing exam that gates someone's career do not need — and should not get — the same level of monitoring.",[34,466,467,468,471,472,474],{},"Proportionate security means asking, for each assessment: what is genuinely at stake, what threats actually apply, and what is the least intrusive control that addresses them? Sometimes that is full ",[41,469,470],{"href":135},"proctoring with lockdown",". Often it is much less — good ",[41,473,115],{"href":114},", randomised question banks, sensible time limits and access controls that make cheating hard without treating every candidate as a suspect.",[52,476,478],{"id":477},"designing-for-trust-as-well-as-integrity","Designing for trust as well as integrity",[34,480,481],{},"A few principles help keep the balance right:",[101,483,484,493,498],{},[104,485,486,488,489,492],{},[108,487,170],{}," collect the minimum data needed for the stakes involved, and be deliberate about what you ",[37,490,491],{},"don't"," collect.",[104,494,495,497],{},[108,496,176],{}," tell candidates what is monitored, why, and what happens to the data. Surveillance people understand and consent to erodes far less trust than surveillance that feels hidden.",[104,499,500,503],{},[108,501,502],{},"Governance, not just tooling:"," decisions that affect someone's result should have a human in the loop, a clear appeals route and an audit trail — not an automated flag standing in for a judgement.",[52,505,507],{"id":506},"the-real-goal","The real goal",[34,509,510],{},"The point of exam security is a result everyone can believe in — the candidate, the awarding body and the people who rely on the credential. Maximum monitoring is a tempting proxy for that, but it is only a proxy, and an expensive one. Integrity, trust and proportionality are not in tension when security is designed well; they reinforce each other. The organisations that get this right are not the ones watching the most. They are the ones watching the right things, for the right reasons, at the right stakes.",[34,512,513],{},[37,514,39,515,47],{},[41,516,411],{"href":409,"rel":517},[45],{"title":195,"searchDepth":196,"depth":196,"links":519},[520,521,526,527,528],{"id":422,"depth":196,"text":423},{"id":429,"depth":196,"text":430,"children":522},[523,524,525],{"id":436,"depth":202,"text":437},{"id":443,"depth":202,"text":444},{"id":453,"depth":202,"text":454},{"id":460,"depth":196,"text":461},{"id":477,"depth":196,"text":478},{"id":506,"depth":196,"text":507},"This article was originally featured on Security Journal UK.",{"url":531,"altText":532},"\u002Fimg\u002Fblog\u002Fintegrity-trust-surveillance.svg","Illustration of a balance scale weighing integrity against surveillance",{},"\u002Fblog\u002Fbalancing-integrity-trust-and-surveillance-in-exams","2026-05-11",{"title":400,"description":529},"Maximum surveillance is not the same as maximum integrity. Why proportionate, risk-based exam security protects trust, fairness and legitimacy better than monitoring everything.","Balancing integrity, trust and surveillance in online exams","balancing-integrity-trust-and-surveillance-in-exams","blog\u002Fbalancing-integrity-trust-and-surveillance-in-exams","More monitoring does not guarantee more integrity, and can carry real trust and legal costs.",[225,226,227,543],"candidate-experience","WIbEfbOcmbbwqhdaDKxo0oFy4CSudhiSLznVXazvrJk",[],1784159169333]