Summer Savings from ACT
Get 26% off courses and products including classroom, virtual, e-learning and books throughout this July.
**EXCLUDES AWARDING BODY FEES, SHIPPING AND EXAMINATION ADMIN COSTS
Get 26% off courses and products including classroom, virtual, e-learning and books throughout this July.
**EXCLUDES AWARDING BODY FEES, SHIPPING AND EXAMINATION ADMIN COSTS

This guide looks at current guidance from NEBOSH about the use of AI in Scenario Based Assessments (SBAs), considers the ethics of using it to support you in your assessment, and answers our most frequently asked questions.
Let’s start with the basics – when we talk about ‘AI’ in this context, we are referring to any online tool that will generate content from a prompt that could then be used in your assessment. This includes chatbots such as:
These programs are what most will think of when they think of AI, however you may not realise that plugins such as Grammarly and certain features in ProWritingAid are also AI-based tools, and are recognised by software that checks for the use of AI in a body of text.
NEBOSH’s stance on the use of AI has changed over time as technology changes and is likely to continue to evolve as they regularly reassess their qualifications, learners and policies.
In December 2025, NEBOSH released the latest version of their Malpractice Policy which stated for the first time that:
“certain uses of Artificial Intelligence (AI) may be permitted in some types of assessments…However, unless specifically permitted to do so (eg by assessment instructions, or any granted reasonable adjustments), the use of AI during an assessment is prohibited and will be considered malpractice” – NEBOSH Malpractice Policy, Version 21 (December 2025)
This means that AI can only be used with explicit and direct permission from NEBOSH themselves – not Learning Partners, tutors, or anyone else that may be involved in your studies. The mention of ‘assessment instructions’ does seem to suggest that NEBOSH may be considering incorporating the use of AI in to certain assessments in the future, however there are currently no NEBOSH qualifications, at any level, that allow the use of AI.
The Malpractice Policy explicitly states that the misuse of AI will still be treated as malpractice and investigated the same as any other violation – for example, sharing answers or copying and pasting information found on the internet. The bottom line is that all work going in to a NEBOSH exam should be your own original thoughts, ideas and analysis.
When you submit an assessment to NEBOSH, you are declaring that it is all your own work. If any part of your answer is found to have been written by AI, then you will be considered in breach of the Malpractice policy – just like if you had asked another human to write an answer for you.
However, there are very narrow circumstances where the use of AI may be allowed. The ‘Guidance to Learners’ section found on the front page of Scenario Based Assessments (SBAs), which are used for NEBOSH Certificate level courses such as the National General Certificate as well as in several units of the NEBOSH Diploma, has recently been updated to provide some guidance on the use of AI:
You may only use AI-enabled assistants or tools for planning your work…But you must not use AI-generated output in any form in your assessment…Your thoughts, ideas, interpretation, analysis and words must be your own. – NEBOSH Unit NG1 Assessment, Feb 2026.
So what does this mean exactly?
For example, let’s say that a task in your exam asks you to comment on how the site manager in the scenario has supported their apprentice in health and safety matters.
In this case, AI is simply acting as a search engine that is providing you with public information, and all of your work is still your own analysis and uses your own judgement based off of your existing knowledge.
You should be mindful that AI cannot be relied on to be 100% correct – you should always factcheck any information that you use to plan your answer, and most importantly, never copy and paste anything from AI in to your answer. Remember that in an SBA, you should always refer your answer back to the scenario provided in order to gain marks. An AI answer will lack the context that you need to gain the mark, and will never be able to provide the insight that you should be able to give after studying the syllabus.
When it comes to malpractice and the misuse of AI, the same rules apply to NEBOSH risk assessments as apply to the examinations.
All of your work must be your own, and any information that’s found to have come wholly or partly from AI will put you in violation of the Malpractice policy. Remember that your risk assessment is specifically for your workplace – AI does not know your workplace!
NEBOSH’s Malpractice policy includes the following when it comes to the use of AI:
“…unless specifically permitted to do so (eg by assessment instructions, or any granted reasonable adjustments), the use of AI during an assessment is prohibited”.
As of March 2026, the Access Arrangements and Reasonable Adjustments Policy doesn’t mention the use of AI. However, if you believe that you have a condition (as defined under the Equality Act 2010) where the use of AI will allow you to complete the assessment on the same level as someone without your condition, you are able to request a reasonable adjustment prior to your assessment. In previous years, when the assessments were handwritten and took place in a classroom, this included extra time to complete or access to a dictionary. Now that the exam is open book, online, and can be completed over a 24-hour period, it’s more accessible than it’s ever been but you are still able to request additional support where needed.
You should be aware that these are granted at NEBOSH’s discretion – they are able to deny or provide other reasonable alternatives if they feel that your suggestions are not reasonable. You will also need to provide evidence to support your request.
Under no circumstances should you use AI to assist with a disability or condition without formally applying to NEBOSH beforehand and being granted permission. Should they grant you any kind of reasonable adjustment then you will still need to apply for any further assessments you take, as it’s not a rolling agreement.
In their Malpractice Policy, NEBOSH provide the following examples as misuse of AI in assessments:
Although it is not explicitly mentioned in NEBOSH’s Malpractice Policy, using AI tools like Grammarly to check grammar and spelling is also forbidden. It’s important to remember that you do not lose marks for spelling and grammar – NEBOSH are looking to test your understanding and application of health and safety, not your literacy skills. As long as your answers can be clearly understood by the examiner, minor spelling mistakes have no effect on your mark.
Your Learning Partner should provide your with a copy of the Malpractice Policy in preparation for your assessments, but you can also download the latest version via the Policies and Procedures section of the NEBOSH website.
NEBOSH combine a range of methods to checking for malpractice – these include your closing interview (or professional discussions for NEBOSH Diploma learners), academic plagiarism software, and metadata analysis.
If NEBOSH suspect that your submission contains AI-generated content then they will open a formal investigation against you in line with their Malpractice policy. This investigation may include requesting a statement from you about the allegations, asking you to submit evidence, and contacting your Learning Partner to request supporting evidence of your studies. You’re unable to sit any more NEBOSH assessments until the investigation is over.
The consequences of being found in breach of the malpractice policy vary from case to case, and will depend on your specific circumstances. These can range from your results being declared void to a lifetime ban. You won’t receive a refund for any exam fees associated with a voided result, and you’ll also have to pay to resit next time – depending on the qualification, you may have to wait a few months for the next available sitting, making a voided result a waste of your time as well as money.
There are also wider implications – NEBOSH qualifications are often a requirement for many health and safety professional roles, and a blanket ban from taking NEBOSH qualifications can be a big red flag for your current and future employers.
If you have any questions about using AI in your NEBOSH assessments, you should contact your Learning Partner directly before taking any action.
The bottom line when it comes to AI use is that the safest and most ethical option is to avoid it altogether. It’s understandable to want to give yourself the best possible chance when it comes to an assessment that could impact your entire career, but the knowledge that you demonstrate during your exam will also be the knowledge that you take in to your career as a health and safety professional. NEBOSH Scenario Based Assessments are designed to test your competency, and if you don’t understand the subject well enough to pass the exam then the unfortunate truth is that you also don’t understand the subject enough to put it in to practice in the real world.
The consequences of cheating or taking shortcuts in health and safety are illness, injury and even fatalities, which is why malpractice at this level is taken so seriously.