AI in e-assessment: why trust, not technology, will define the next generation of Ed tech
Artificial intelligence is now embedded across education technology. It can generate assessment content, mark work, monitor exams and support learners, teachers and examiners. At the same time, AI tools are available to virtually every student, raising fundamental questions about exam integrity and the future of assessment.
ECI Partners, the leading mid-market investment firm, recently invested in Paragin Group, the Benelux market leader in high-stakes digital assessment software, and for investment team partner, Rory Nath, one message has become increasingly clear: AI alone will not determine which organisations succeed; the real differentiator will be trust. He comments: “This year’s e-Assessment Association conference captured the challenge perfectly with one of the session titles: Guilty until proven human. It reflected the growing concern across the education sector about maintaining confidence in qualifications in an age where AI is becoming ubiquitous. A week later, on a panel at EdTechX, Nath heard investors asking the same question from the other side: “With AI changing everything, which of these businesses are actually backable?”
Here are Nath’s thoughts on the factors that will define the next generation of edtech:
Innovation needs more than AI
Today, almost every edtech supplier is demonstrating AI-powered features, from question generation and automated marking and feedback, to identifying potentially suspicious behaviour during exams. While these capabilities are valuable if implemented well, they are increasingly built on the same underlying large language models available to everyone. If a product is just a thin layer over a model that a competitor can license tomorrow, it does not offer sustainable differentiation.
The organisations creating the greatest value are those combining AI with deep subject expertise, proprietary content and a detailed understanding of educators’ workflows. A good example is UK-based medical education provider Risr, which has worked with the Royal College of General Practitioners (RCGP) to develop an AI-powered patient and coaching system for trainee GPs preparing for clinical assessments. The effectiveness of the platform comes not simply from AI itself, but from the specialist clinical knowledge and assessment content behind it.
The same principle applies across education. AI becomes significantly more valuable when it sits on layers of vertical depth such as proprietary content or deep knowledge of complex customer workflows.
Paragin’s own approach reflects this philosophy. Having spent more than 20 years working with schools, awarding organisations and governments, the company has developed one of the market’s most comprehensive assessment platforms. AI is helping accelerate product development, but the direction continues to be shaped by deep understanding of assessment practice and customer needs. It is this last 10% of depth in customer knowledge that impressed the ECI team most about Paragin, and it is this human expertise that gives it the edge in its tech roadmap.
For Jeroen Bakker, CEO of Paragin Group, staying front footed on this topic is important: “I think AI needs to find a place in the software layer that adds real value to end users. Valuable use of AI arises when it actually helps in developing and assessing knowledge and competencies, rather than merely adding AI for AI’s sake. Reliability, transparency and explainability are essential: it would be very bad for an assessment to be inexplicable or to unintentionally differ per student, or for the result to be slightly different tomorrow than today, because the LLM received an update last night.”
He adds: “It is easy to add AI and feel you have found a solution, but does it really do what we want it to do, and do we have the right guardrails in place to control what goes in and what comes out? Those questions guide our approach at Paragin and we are very considered in how we think about AI, so that we continue to invest and innovate in a way that actually delivers value to our users.”
Trust is becoming the defining competitive advantage
Strip away the technology and what an awarding body ultimately sells is confidence. Schools, universities, employers and professional bodies need to know that qualifications genuinely reflect a learner’s knowledge and abilities. This is not an environment with much room for experimentation: hundreds of thousands of students can sit a single exam, such as a GCSE, on the same morning, and any error is felt immediately and at scale. That is why the market will be fairly conservative on AI adoption and reliability, fairness and transparency are likely to matter far more than simply being first to introduce new features. Slower but deliberate is perhaps a better description of what is meant by conservative here. This also explains why established providers continue to hold strong positions in the market. While AI capabilities can often be replicated quickly, trust built over many years cannot.
Security and fairness must be integral
As AI becomes more prevalent in e-assessment, expectations around governance are increasing. Assessment platforms handle highly sensitive personal data, and the results themselves can decide whether someone qualifies as a doctor, an accountant or an engineer, so security, privacy and fairness need to be embedded from first principles throughout product design, rather than added later.
Regulation is increasingly reflecting these expectations and under the EU AI Act, AI used in educational assessment and candidate monitoring is classified as high risk, requiring robust governance, human oversight and evidence that systems operate fairly. Although the original August 2026 implementation deadline has now been pushed back to 2 December 2027 under the Digital Omnibus package, the direction of travel remains clear, and similar expectations are emerging internationally.
For organisations that have spent years building secure and trusted assessment systems, these developments reinforce the importance of long-term investment in compliance, governance and product quality. Paragin, for example has spent over 20 years wrestling with exactly these questions while continuing to innovate, and it has the scale to treat compliance and security as core to the product rather than a tax on growth. As the bar on security and compliance continues to rise, it is likely to become harder for smaller, less-resourced providers to keep pace on their own, which points to further consolidation across the sector in the years ahead.
Looking beyond the AI race
The long-term winners in e-assessment will likely not be those with the most AI features, instead, it will depend on who has earned the confidence of learners, educators, regulators and employers.
Nath’s view is that the winners over the next three to five years will be the ones that treat integrity and compliance as the point of the exercise rather than the price of it, because that is exactly what AI cannot commoditise.