Artificial intelligence, assessment and patient safety
I recently marked a Level 7 assignment that genuinely made me stop and think.
After almost 30 years in nursing and a good few of them in healthcare education, you develop an instinct when reading postgraduate work. It is difficult to explain, but experienced educators quickly recognise the difference between a student who understands a subject and one who has simply found the right words. It isn’t about grammar, presentation or academic language. It is about the thought process behind the submission.
This particular assignment was well written. Every question had been answered. The spelling was accurate, the grammar was excellent and, at first glance, it looked like a strong piece of work. Yet something didn’t feel right.
The assignment answered the questions, but it didn’t explore them. There was very little critical analysis, limited evaluation of the evidence and almost no indication that the student had weighed up different viewpoints before reaching a conclusion. Rather than demonstrating postgraduate thinking, it felt as though it was presenting information.
As educators, that difference matters enormously.
As part of our quality assurance process, we reviewed the submission using an AI detection tool. The report suggested that a high proportion of the content may have been generated using Artificial Intelligence.
Let me be absolutely clear. That result did not prove misconduct. AI detection software is still evolving and should never be used as the sole basis for an academic judgement. False positives and false negatives are well recognised, and no responsible educator should rely on a percentage score alone. However, the software didn’t create our concern. It simply reinforced the questions that had already arisen while reading the assignment.
That experience prompted me to ask a much bigger question.
In the age of Artificial Intelligence, how do we know that a postgraduate qualification genuinely represents the knowledge, judgement and clinical reasoning of the person who receives it?
For me, that is now one of the most important questions facing aesthetic education.
This is not an argument against AI
Before anyone misunderstands my position, let me say this. I use Artificial Intelligence.
Like many educators, I have found it incredibly useful. It can summarise research, help structure ideas, improve written communication and explain complex concepts in different ways. Used ethically, it has enormous potential to support learning and lifelong professional development.
The problem isn’t Artificial Intelligence. The problem is how we assess learning in an era where AI can produce work that appears academically polished.
Education has always evolved alongside technology. We moved from handwritten assignments to word processors. We embraced online journals, virtual learning environments and digital libraries. Each development made education more accessible. Artificial Intelligence is simply the next stage in that journey. The challenge is ensuring that our assessment methods evolve at the same pace.
What does Level 7 actually assess?
One of the biggest misconceptions I encounter is that Level 7 simply means a harder assignment. It doesn’t.
A Level 7 qualification should demonstrate far more than the ability to answer questions correctly.
At postgraduate level, students are expected to analyse evidence, critically evaluate published literature, compare contrasting viewpoints, justify clinical decisions and reflect upon their own practice. They should be able to explain why they have reached a particular conclusion, not simply state what that conclusion is.
In aesthetic medicine, that distinction is vital. We are not educating people to pass assignments. We are educating practitioners who will make independent clinical decisions affecting real patients.
Can they recognise the early signs of a vascular occlusion?
Can they identify unrealistic patient expectations?
Can they justify delaying treatment when a patient insists on proceeding?
Can they explain why one intervention is appropriate while another is not?
These questions require judgement, not simply knowledge. That is what Level 7 should be measuring.
Patient safety starts in the classroom
When people think about patient safety, they often think about the treatment room. I think it begins much earlier. It begins during education.
Every assessment we design should ultimately answer one simple question: Would I trust this individual to treat a member of my own family?
If the answer is uncertain, then we owe it to both the student and future patients to investigate why.
Patients assume that qualifications represent competence. They believe that somebody has rigorously assessed the practitioner sitting in front of them. They rarely ask how those assessments were completed. Perhaps, in the future, they will.
If educators cannot confidently demonstrate that a qualification reflects genuine understanding and clinical reasoning, then public confidence in education itself may begin to weaken.
That is a conversation none of us should ignore.
The licensing scheme makes this even more important
The timing of this discussion could not be more significant.
The Government has already set out its intention to introduce a licensing scheme for non-surgical cosmetic procedures in England, with competence, education and patient safety expected to play central roles in future regulation.
As an educator, I welcome higher standards. I believe our profession needs them. However, stronger regulation also places greater responsibility on education providers. If qualifications become part of the evidence that practitioners use to demonstrate competence, then those qualifications must genuinely reflect independent knowledge and clinical judgement.
Anything less risks undermining the very purpose of regulation. Licensing should reassure the public. It should not simply reassure paperwork.
Perhaps assessment needs to change
I don’t believe the answer is to ban Artificial Intelligence. Nor do I believe we should return completely to the educational systems of 30 years ago. Online learning has transformed education for many healthcare professionals. It has increased flexibility, widened access and supported lifelong learning. Those are positive developments.
What I do believe is that assessment deserves another look. Perhaps postgraduate assignments should be supported by structured viva examinations where students explain and defend their work.
Perhaps unseen written examinations completed under supervised conditions should become part of Level 7 programmes again.
Perhaps greater emphasis should be placed on case-based discussions, observed clinical reasoning and Objective Structured Clinical Examinations (OSCEs).
A student who genuinely understands their subject should have no difficulty discussing it with an experienced assessor.
In many cases, a thirty-minute professional discussion tells you far more than thirty pages of beautifully written text.
AI belongs in education – but with boundaries
Ironically, I think Artificial Intelligence should become part of the curriculum. Students should learn how to use it responsibly. They should understand where it can add value and where it has limitations. They should know how to verify information rather than simply accepting it. Most importantly, they should understand that AI can assist learning, but it cannot replace independent professional judgement.
There is a significant difference between using AI to improve grammar or organise ideas and asking it to produce an entire academic submission.
Educational providers need clear guidance on where that boundary lies. Equally, students need confidence that using AI ethically is acceptable, provided they remain the author of their own thinking. Transparency will become increasingly important.
The human element must not be lost
Some of the most valuable learning I have witnessed has never appeared in a written assignment. It has happened during conversations. It has happened when a student has changed their opinion after hearing a different perspective. It has happened when an experienced clinician has openly discussed a complication and explained what they learned from it. Those moments shape judgement. They build confidence. More importantly, they produce safer practitioners.
Artificial Intelligence can generate information. It cannot replace experience. It cannot recognise uncertainty in a patient’s voice. It cannot sense when something simply doesn’t feel right during a consultation. Clinical judgement is developed through reflection, discussion and experience. Those remain uniquely human qualities.
Looking ahead
Artificial Intelligence is not going away. Neither should it. The question is not whether we should use AI in education. The question is whether our methods of assessment are keeping pace with the technology available to our students. That is a challenge for every educator, awarding organisation and training provider.
As aesthetic medicine moves towards greater regulation and licensing, assessment integrity will become just as important as curriculum design. Our responsibility is not simply to award qualifications.
Our responsibility is to ensure that those qualifications genuinely represent the competence, judgement and professionalism expected of those who hold them.
Patients place enormous trust in aesthetic practitioners. They assume that education providers have done everything possible to ensure that the certificate on the clinic wall represents real knowledge, real clinical reasoning and real competence.
That trust should never be taken for granted. Artificial Intelligence may be capable of producing an impressive assignment. It may even help someone structure their ideas more effectively. But it cannot recognise a vascular occlusion. It cannot manage an anaphylactic reaction. It cannot make a difficult ethical decision during a consultation. Only a competent clinician can do that.
If the future licensing scheme is truly about improving standards and protecting patients, then perhaps the most important question is not what qualification a practitioner holds.
Perhaps it is this: How do we know that the qualification genuinely represents the person who earned it?
For me, that is the conversation aesthetic education should be having today – before technology moves faster than our ability to assess the clinicians of tomorrow.
For more information please visit https://nataonline.co.uk/