Fraud, Plagiarism and AI in Admissions: Are Regulatory Bodies Really Equipped to Detect New Forms of Cheating?

A report published in January 2026 by the Commissioner for Admission to Professions in Quebec, under the Office des professions du Québec, confirms what many had suspected but lacked the data to prove: artificial intelligence has become a vehicle for fraud and plagiarism in the admission processes of regulated professions.

For Nexam, this report resonates with particular significance. It validates a reality we have been observing in the field for several years and raises an urgent question for regulatory bodies and associations: are we equipped to respond?

And while this finding is documented in the context of regulated profession admissions, it would be naive to think it ends there. By extension, we can reasonably expect to find the same reality in every sphere where assessment plays a determining role: academic and university settings, hiring and recruitment processes, professional certifications, and entrance examinations. Wherever an assessment determines an important decision, the risk of inappropriate AI use exists.

Is AI Being Used to Cheat in Evaluation and Admission Processes?

The systematic review conducted across 46 Quebec regulatory bodies documents for the first time in Quebec the inappropriate use of AI at three key stages of the admission process:

• Examinations: one order reported the use of AI tools during an examination session
• Portfolios and work samples: AI-based plagiarism documented, including the copying of automatically generated content
• Internships: inappropriate use of AI in written work produced during an internship

This is no longer a working hypothesis. It is a measured reality, recorded in an official Quebec government document.

And this is only the tip of the iceberg. The report itself notes that the number of detected cases remains “relatively low” — which does not mean such situations are rare, but rather that few orders currently have the mechanisms in place to identify and document them.

Why Are Most Regulatory Bodies Unable to Handle a Fraud Case Before a Licence Is Issued?

The report highlights a structural weakness: 58.7% of regulatory bodies have no regulatory framework to handle fraud or plagiarism before a licence to practise is issued. Furthermore, 67.4% have no formalized investigation procedure when a case is suspected.

In other words, even when fraud is detected, many orders do not know exactly how to respond or on what legal basis to act. And when regulatory frameworks struggle to keep up, assessment tools themselves can play a decisive role in prevention.

How Can the Use of AI in Professional Assessments Be Detected?

Generative AI has profoundly changed the nature of plagiarism. It is no longer just about copying and pasting text found online. Today, a candidate can produce in seconds a text that appears original but is entirely machine-generated — indistinguishable from authentic work to the human eye.

Faced with this reality, regulatory bodies and associations need three essential capabilities:

1. Detect the use of AI in written submissions and evaluations
2. Document cases rigorously to be able to act on solid grounds
3. Intervene in a fair, transparent and defensible manner

It is precisely on detection and documentation that solutions like those developed by Nexam make the most sense. Integrating monitoring and detection tools directly into the assessment process — whether for online exams, submitted work or portfolios — helps bridge the gap between the reality on the ground and organizations’ capacity to intervene.

What Can Be Done Concretely to Protect the Integrity of the Admission Process Before a Licence Is Issued?

The Commissioner’s report recommends granting orders a generalized and formalized authority to intervene in fraud situations before a licence is even issued. This is an important legislative avenue, but one that will take time to materialize.

While waiting for the law to evolve, orders are not without recourse. In parallel with regulatory reforms, concrete measures can be implemented right now to strengthen the integrity of the admission process:

• Have candidates sign an integrity declaration at each key stage
• Clearly communicate the rules on AI use before each assessment
• Design authentic examinations that minimize the use of AI through situational scenarios, practical cases or professional judgment questions
• Integrate video or audio questions into assessments, given that cheating is significantly harder in oral formats than in written ones
• Implement active exam monitoring, whether remote or in person
• Use tools that generate documented evidence in cases of suspected misconduct
• Maintain a complete audit trail of assessments to be able to act on solid grounds if fraud is detected, whether before or after the licence is issued

The law can define sanctions. It cannot detect fraud on behalf of organizations. That is why acting now, with the means available, is not only possible — it is necessary.

What Is the Next Step for Professional Orders in the Face of Rising AI Fraud?

The report from Quebec’s Commissioner for Admission to Professions is a clear signal. AI in assessment processes is no longer a future risk to anticipate: it is a present challenge to address.

For regulatory bodies, associations and all organizations managing admission processes, the question is no longer whether they need to act, but how and with what tools.

Source: Commissioner for Admission to Professions in Quebec, Portrait de l’admission aux professions: Fraud and Plagiarism in the Admission Process for Regulated Professions, (available in French only), January 2026.