It’s been just over a month since the ATP Innovations in Testing conference in New Orleans, and the energy remains high. While the beignets were sweet and the therapy puppies provided much-needed dopamine hits, the real takeaway was the shift in atmosphere. For the nearly 1,100 attendees and across 150+ sessions, the industry seems to have moved from a state of AI panic to one of AI acceptance within assessment.
At risr/, as a company led by innovation, we leaned heavily into this transition. We had the privilege of participating in several sessions this year, focusing on how we bridge the gap between legacy testing and future-proof tech. We also showcased our award-winning AI Clinical Coach, our vision for the future of simulated learning and assessment, which directly addresses the industry’s need for more authentic, performance-based evaluation.
The Crossroads: Guilty Until Proven Human
One of the most recurring themes was the strain on test security. I’ve been using the phrase “Guilty until Proven Human” to describe the current state of high-stakes testing. We are at a crossroads where the defensive “wall-building” approach is meeting its limit.
The data backs this up. In one Test Security session, attendees were asked: “How do you feel about the threat posed by AI in 2026 compared to 2025?” 73% said it was more of a threat while 24% said it was about the same and only 3% felt the threat had lessened.
This anxiety highlights a growing digital divide. As some students gain access to cutting-edge Agentic AI tools while others lag behind, the equity of access becomes a moral and logistical minefield. The Test Centre vs. Remote Proctoring debate has also resurfaced, epitomized by recent headlines from the UK on a step-backwards from Remote Proctoring for a large awarding body, proving that we haven’t yet found a silver bullet for global security. Maybe controversially, I don’t think the silver bullet exists…
The Shift Toward Authentic Assessment
At risr/, we believe the solution isn’t just better proctoring; its better assessment design. This is where Authentic Assessment and Agentic AI come into play.
- AI: The Future is Human: Back in 2024, a panel we led suggested that the “human element” would be the ultimate differentiator. At the time, that was not a commonly shared opinion, as it was felt AI would replace everything. Today, that sentiment has somewhat flipped. Since everyone now has access to the same LLMs, the true value lies in your Subject Matter Experts (SMEs) and how they curate and validate AI outputs (and inputs).
- Guardians vs. Adversaries: We need to decide if AI is a tool to be locked out or a guardian to be integrated. Used in the correct manner, AI is a hugely valuable co-worker, tutor, coach, critical friend and a significant value addition to the assessment process.
In health professional assessment, we must ask: Is our goal to exclude AI, or to evaluate how effectively a practitioner collaborates with it for better patient outcomes?
Three Strategic Shifts
During one talk, three strategic shifts for organizations currently building walls around a moving target were proposed:
- Move from Detection to Design: Stop trying to “catch” AI and start designing tasks where AI is either a baseline tool or irrelevant to the core competency.
- Global Problems, Local Solutions: While AI is global, regulation and legislation are local. We must build flexible architectures that respect regional data privacy and ethical standards.
- Collaboration over Exclusion: In the real world, professionals use tools. Our assessments should reflect that reality, moving toward “open-resource” models that test judgment rather than recall.
Final Thoughts: A Community of Practice
Ultimately, ATP is about the “three Cs”: Connections, Collaboration, and Community. We are all grappling with the same fundamental question: are we fighting the wrong battle? If we spend all our resources on test security, we might miss the opportunity to actually improve the validity of our exams.
The question is not whether AI will keep changing the threat landscape. It will. The question is whether our assessment architectures are designed to flex with that change, or if we are still building static walls around a target that has already moved.
We will succeed in this rapidly changing landscape, if we continue to work together, share experiences and dare I say, try to be different? Challenge the norms, don’t just put ‘paper behind glass’ and think you are innovative. Push the boundaries of what is possible and whether we like it or not (I do!) AI is a huge part of our future and can revolutionize assessment and learning worldwide.
Where are we next? The conversation continues! Catch the risr/ team at these upcoming events:
- ICAM (International Congress on Academic Medicine)
- AACOM (American Association of Colleges of Osteopathic Medicine)
- CNG (Certification Network Group)
- eAA (e-Assessment Association)
- ANZAHPE (Australia & NZ Association for Health Professional Educators)
- AMEE (International Association for Health Professions Education)
- e-ATP
Whether you are rethinking how clinical competency gets assessed, looking for ways to give your learners more equitable access to preparation, or simply want to explore what AI-powered simulation looks like in practice, we would love to talk.
You can get a trial of AI Clinical Coach, or get in touch with the team to start the conversation.
