It's one of the most common questions students ask about case study revision. You've written an answer to a practice task, you have no one to mark it, and ChatGPT is open in the next tab. Can it tell you what you'd have scored?
The short answer: a general chatbot can give you useful comments on an answer, but not a mark you should rely on. This article explains why, how CIMA's markers actually reach a mark, and how to use a chatbot in a way that helps your revision rather than misleading it.
What a chatbot does well
Used carefully, a general chatbot is a good revision partner for some jobs:
- Explaining a technical topic in different words when your study text hasn't made it click, such as the difference between a sales mix and a sales quantity variance.
- Challenging your reasoning. Ask it to find each place where you've asserted something without saying how or why.
- Checking you answered the question. Paste the requirement and your answer, and ask whether every part of the requirement is covered.
- Quizzing you on your own pre-seen notes, so you can test your recall.
All of these use the chatbot to question your work, not to judge it. That difference matters, for four reasons.
Why it can't mark your answer
It doesn't have the marking guidance
CIMA's markers work from a marking scheme written for each variant of the exam. It breaks every sub-task into traits and says what Level 1, Level 2 and Level 3 look like for each one. A chatbot has none of this. When you ask it for a mark, it makes up its own scheme on the spot, and you can't see what that scheme is.
Its levels aren't consistent
Ask the same chatbot to mark the same answer twice, in two new chats, and you will often get two different marks. Change how you word the request and the mark moves again. Chatbots also tend to be encouraging and to go along with how you frame things, so "is this a pass?" can get a kinder answer than "what's wrong with this?". A mark that changes when nothing in your answer has changed can't tell you whether you're improving.
It doesn't know your pre-seen
The case study is marked on how well you use the information you're given and apply it to the company. Unless you paste the pre-seen in, the chatbot hasn't read it. If it searches the web, it may find someone else's summary, or a pre-seen from an earlier sitting. It then fills the gaps with facts that sound right. An answer can look well applied to a chatbot that doesn't know the company, and generic to a marker who does.
It invents exam facts
Chatbots answer questions about the exam with confidence, and can be out of date or simply wrong. Common examples:
- that answers are marked on four competencies (technical, business, people and leadership). That was the earlier syllabus's model, not how CIMA marks the case study exams now;
- that you need a minimum in each core activity to pass. There is none: the pass or fail is for the exam as a whole;
- a raw pass mark, often quoted as 50% or 80%. The pass is a scaled score of 80 out of 150, and CIMA doesn't publish how raw marks convert to it.
Check anything a chatbot tells you about the exam against CIMA's own documents, listed under Sources below.
How CIMA actually marks
The case study exams are marked by trained human markers, with standardisation and moderation behind them. In short:
- Each sub-task is split into one or more traits: a topic, a counted item such as "second recommendation", or one side of an argument.
- Each trait is placed at Level 1, 2 or 3 on a best-fit basis, with its marks spread roughly evenly across the three levels. An answer with nothing worth rewarding on a trait scores 0.
- The level is judged on four elements: technical understanding, communication, use of the information, and application to the scenario. At Management and Strategic level, communication means answering what the task actually asks.
- Marking is positive. You're rewarded for what you show, not penalised for what you leave out, and other valid approaches earn credit.
- An answer that doesn't address the requirement earns no marks, however good it is in itself.
The examiner reports describe what separates the levels: following the verb, explaining how and why, applying knowledge to this company rather than reciting it, and developing a few points in depth. We cover all of this in how case study answers are marked.
Using a chatbot safely for revision
If you use a chatbot, use it to sharpen your own judgement, not to replace it.
- Ask for questions, not a mark. "Which of my points would a manager ask 'so what?' about?" is more useful than "what would this score?".
- Give it the requirement. Most low marks come from answering a different question. Ask it to check your answer against each part of the requirement and the verb.
- Check technical content against your study text. If the chatbot explains a treatment or a calculation, confirm it in a source you trust before you learn it.
- Don't learn answers it writes. The examiner reports say plainly that rote-learned answers score very few marks, if any, and that all the unseen information in a task is there for a reason.
- Use CIMA's post-exam materials as your marking guide. For past variants, CIMA publishes the tasks, suggested answers, marking guidance and examiner reports. Mark your own answer to a past task against its published marking guidance: no chatbot can give you a better reference.
- Look after your data. Check the chatbot's settings on whether your chats are used to train its models. CIMA's pre-seens and exam materials are copyright, so think before uploading whole documents to any tool.
Sources
CIMA's documents, read in October 2026:
- Operational case study pilot paper pack, from May 2026: marking guidance, the four elements and candidate guidance.
- Operational case study May and August 2026 post-exam materials: the examiner's report on rote-learned answers and generic points.
- Management case study May and August 2026 post-exam materials: the four elements at Management level, and the examiner's report.
- A guide to your examination results, revised November 2025: the scaled score and the pass at 80.
- The case study exam marking process, AICPA & CIMA.