Across the world, assessment is moving from fixed physical delivery to hybrid and remote models. It is a shift being driven by demand for greater choice, flexibility and equity, and one that is unlikely to reverse. Better technology and stronger security mean that greater flexibility does not have to come at the expense of confidence in the outcome.
At PSI, we invest across both physical and remote testing so customers can choose the right model without compromising security.
While the assessing institution or organisation is our customer, the candidate experience always sits at the centre of the model. Wherever possible, we want organisations to be location agnostic in their student, workforce or immigration assessment, breaking down barriers created by geography, mobility and budget.
Candidates have different needs and circumstances. For someone living in a busy household or dealing with poor connectivity, a test centre may provide the best experience. For others, particularly those who cannot easily travel to a specific site, remote delivery can make an assessment significantly more accessible.
Ultimately, the delivery model matters less than the security architecture and controls that surround the assessment.
At PSI we operate, and will continue to operate, test centres in around 140 countries worldwide. But we have invested heavily in our remote testing model to ensure institutions and candidates can be sure of equivalent security standards in both models.
Across more than 2 million completed tests, we have seen the difference in outcomes between online and in-person testing narrow from more than 11% to less than 5%.
Security is layered
Our work starts with knowing that the right person is taking a test. Identity assurance can now go far beyond checking a document at the beginning of an assessment. Face, voice and keystroke biometrics help establish that a candidate matches their registered identity, while behavioural indicators identify potential impersonation, and that assurance continues throughout the assessment. With these connected identity controls, we have seen a 160%+ increase in confirmed impersonation cases.
For remotely proctored assessments, technology also gives us a much richer understanding of the candidate’s testing environment. Second-camera monitoring extends visibility beyond a traditional webcam, helping identify external assistance, hidden devices and activity outside its field of view. Around 35% of confirmed violations we identify involve fraud detectable only through that second camera.
Behavioural monitoring adds another layer, with AI-assisted technology identifying unusual activity in real time. But this is not about replacing people with algorithms. Technology flags what looks unusual, a trained proctor makes the call. And we’ve seen seven times more real-time security violations confirmed through AI-assisted flagging and human review.
Flexible, but consistent
For me, the strongest reason to keep investing in security is not simply to catch fraud. It is to protect genuine candidates.
Whether someone is applying for a visa, sitting a university assessment, earning a professional certification or demonstrating competence for an employer, they have invested time, effort and often money in that achievement. Every fraudulent result weakens the value of a legitimate one.
Universities, employers, professional bodies and governments need confidence that an assessment reflects the knowledge, skills or abilities it is designed to measure, regardless of how it is delivered
The same trust matters for the organisations relying on those results. Universities, employers, professional bodies and governments need confidence that an assessment reflects the knowledge, skills or abilities it is designed to measure, regardless of how it is delivered.
Different programmes will need different delivery models. What should not vary is the standard. Remote or in person, the result must be secure, fair and able to stand up to scrutiny.
Constantly evolving to meet new demands
There is no finish line. Fraud changes, technology changes, and AI is raising the stakes on both sides. Staying ahead of that takes dedicated attention. PSI’s Test Security Office, a team of over 70 specialists, works to track known threats and identify emerging ones.
That is why security must be built into every stage of the assessment lifecycle, from identity verification to delivery, monitoring and post review.
Modern assessment should give organisations choice without weakening trust. It should give candidates greater access without diminishing the value of what they achieve.
The test may be remote, in person or something in between. What matters is whether the result stands up to scrutiny.
About the author: Wallace Dalrymple is chief security officer at PSI, where he leads a global team of 70 focused on test security and fraud mitigation globally and is accountable for the Office of Test Integrity. His remit covers the full integrity stack: identity assurance, behavioural and forensic analytics, content protection, and secure delivery across remote proctoring and physical test centres for high-stakes exams, licensure, and government testing programs. With more than 30 years in security, Wallace is known for building detection capability that holds up under regulatory, client, and legal scrutiny.
