The Complete Guide to the NEBOSH Malpractice Policy Updates: Navigating AI in Your Exams
The Complete Guide to the NEBOSH Malpractice Policy Updates: Navigating AI in Your Exams
By Will Taylor, Ex-NEBOSH Examiner, CMIOSH, and Lead Tutor at Compassa
If you are currently studying for a NEBOSH qualification, or are about to sit your exams, you are navigating a landscape that has shifted dramatically over the last year. The rise of Artificial Intelligence (AI) tools like ChatGPT, Claude, and Gemini has forced educational bodies worldwide to adapt, and NEBOSH is no exception.
Historically, NEBOSH maintained a zero-tolerance policy regarding AI. However, recent updates to the NEBOSH Malpractice Policy have introduced some of the biggest changes we have seen in the last five years.
As an ex-NEBOSH examiner and the lead tutor at Compassa, I have seen first-hand how these rules are applied, how students accidentally fall foul of them, and how rigorously NEBOSH investigates suspected cheating.
This guide will break down exactly what the new NEBOSH malpractice rules mean for you, what constitutes cheating, and how you can ensure your hard work isn’t voided by a misunderstanding of the rules.
The Big Shift: Is AI Now Allowed in NEBOSH Exams?
Up until December 2025, learners were strictly prohibited from using AI under practically any circumstances. However, a close reading of the updated NEBOSH Malpractice Policy reveals a groundbreaking sentence:
“Certain uses of artificial intelligence may be permitted in some types of assessments.”
At first glance, this sounds like the floodgates have opened. They have not. The policy immediately clarifies that permitted uses will be stated only in the specific assessment instructions for your exact exam. Unless specifically permitted to do so by your assessment instructions, the use of AI during an assessment remains completely prohibited.
If your exam paper does not explicitly state that AI is allowed, using it is malpractice. Always read the “Guidance to Learners” on the front of your specific exam paper before you begin.
The Strict Red Lines: What Constitutes NEBOSH Malpractice?
When learners panic or struggle with time management, the temptation to use AI can be strong. However, NEBOSH has explicitly outlined several red lines. Crossing any of these will trigger a malpractice investigation.
1. Uploading NEBOSH Copyrighted Materials
You cannot take your exam paper, syllabus guides, or course materials and upload them to an AI tool. For example, copying and pasting “Scenario A” into an AI prompt and asking it to explain the hazards is a direct violation of copyright and constitutes malpractice. Just the act of uploading the copyrighted exam paper is sufficient to warrant an investigation, regardless of whether you use the output.
2. Copying or Paraphrasing AI Responses
A common myth circulating on internet forums is that you can have AI write an answer, and as long as you rewrite it “in your own words,” you are safe. This is fundamentally false under the new policy.
NEBOSH mandates that learners must demonstrate their own thoughts, ideas, interpretations, and analysis. If an AI tool tells you which health and safety points to make, and you simply rephrase those points, the underlying thoughts and analysis belong to the AI, not you. Submitting this—whether in part or in whole—is malpractice.
3. Using AI for Calculations
For advanced qualifications, such as the NEBOSH Diploma, you may be required to perform calculations. Using AI to work out these formulas or solve mathematical problems on your behalf means the work does not reflect your own evaluation. You must perform all calculations yourself.
4. Using AI as a Translation Device
This is an absolute killer for many international students. Often, students whose first language is not English feel they can better express their health and safety knowledge in their native tongue. They write their answers out and then use an AI tool to translate the text into English for submission.
Under the updated policy, using AI tools to rewrite a qualification assessment into a different language is strictly classified as malpractice. You must write and submit your answers in the official language of the assessment.
The Exception: Reasonable Adjustments
NEBOSH is committed to accessibility. There are cases where learners with learning disadvantages—such as dyslexia or other cognitive difficulties—might be permitted to use AI-enabled assistive technologies for typing, grammar correction, or proofreading.
However, this is not a blanket rule you can apply yourself. You must formally apply for “Reasonable Adjustments” through your learning partner (your course provider) before the exam. NEBOSH reviews these requests on a case-by-case basis and will provide written authorisation if approved.
How to Legally Use AI (If Permitted by Your Exam)
So, what does it look like when AI is allowed? During the February IG1 exams, NEBOSH updated the learner instructions to allow very restricted use of AI. The instructions stated:
“You may only use AI-enabled assistance or tools for planning your work… for example, to help you find and summarise relevant information sources and to simplify concepts or ideas.”
Even when permitted, the usage is strictly confined to the research and planning phase. Here is how you can use it safely:
- Finding Information Sources: If you have a question about employee consultation, you can ask AI, “What are the official HSE or ILO guidance documents regarding involving employees in the workplace?”
- Simplifying Concepts: If you need to reference HSG245 for investigating accidents, you can ask AI, “Explain how to calculate the levels of investigation according to HSG245.” The Golden Rule: Notice that in these examples, we are not feeding the AI the exam scenario, nor are we asking it to answer the exam question. We are treating the AI like a highly efficient search engine to find guidance documents or summarise public information before we sit down to write our own original thoughts based on the scenario.
Even when AI is allowed for planning, the instructions remain clear: You must not use AI-generated output in any form in your assessment. No copying, no pasting, no paraphrasing.
The Investigation Process: Detection and Consequences
NEBOSH use a combination of technology and interviews to detect malpractice. Consequences include being banned from NEBOSH assessments or having your result “voided”.
If you are tempted to risk it, you need to understand how rigorously NEBOSH enforces these rules. NEBOSH, alongside examiners and tutors, is actively hunting for malpractice using a combination of sophisticated detection software, metadata analysis, and expert human intuition.
Let’s look at the data from the NEBOSH Ethical Practice Report covering just three months (July to September 2025):
- NEBOSH conducted nearly 2,000 investigations into suspected malpractice.
- In my experience reviewing these statistics, roughly 95% of those investigated are found guilty.
- In that single quarter, 463 people were banned from taking NEBOSH qualifications.
- 18 individuals received lifetime bans.
If you are caught, you don’t just get a zero. Your paper is “voided.” This means you lose your exam fee, you must wait months to re-register for the next available exam window, pay the fees again, and go through the immense stress of resitting the paper.
Why AI Fails at NEBOSH Exams Anyway
Aside from the ethical and punitive risks, AI is genuinely bad at passing NEBOSH exams. NEBOSH open-book exams test your ability to apply health and safety concepts to a highly specific, nuanced scenario.
In my experience running AI outputs through our tutor grading systems, AI frequently misunderstands the core of the question. It will provide a plausible-sounding, well-written answer that is entirely disconnected from the specific context of the scenario provided. You might get a beautifully written essay on generic leadership qualities, but score zero marks because you failed to identify the negative leadership behaviors exhibited by the specific manager in the exam brief.
The Ethical Bottom Line
Ultimately, the strictness of the NEBOSH Malpractice policy comes down to the core purpose of our profession.
Cheats do not have a place in the health and safety industry. If you manage to cheat your way to a qualification, you lack the core competency required to do the job. When you eventually secure a role as a health and safety professional, your lack of genuine knowledge will result in bad advice. In our industry, bad advice leads directly to injury, ill health, and fatalities.
If you are struggling with your studies to the point where you feel you need AI to pass, the solution isn’t to cheat. The solution is to seek better training, request tutor support, and engage with materials that actually help you understand the concepts.
Study hard, do the work, and earn your qualification the right way. Your future colleagues—and their safety—depend on it.
Will Taylor is a former NEBOSH Examiner, a Chartered Member of IOSH, and the lead tutor at Compassa. Compassa aims to be the premier provider of the NEBOSH National and International General Certificate, delivering world-class training through innovative interactive video technology with an 84% exam pass rate over the last 12 months.
Leave a Reply