Focus on doing something which gives you the best chance of getting a good grade. If you can do that with something related to AI and Machine learning, then great, but that should be secondary. You want to make sure that whatever you choose isn't going to be so complicated that it drains hundreds of hours and ends up being unrealistic.
A lot of people often end up doing something like a web or desktop app, maybe based on something they already use, or related to studying, or maybe related to a club they're involved in, or their job, etc. for example:
- writing an app which based on some app on their phone such as a whatsapp clone, or a social media site
- quiz/learning/timetable app based on studying
- 'management' app for a club, business/shop, college, etc
- A game of some kind
Whatever you choose, you're better off starting with something which might be simple and then building complexity on top of that idea later on once you've gotten the simplest possible thing working. Anything can be made more complex by adding more features into it. For example:
- Splitting an app between client/server. (Maybe writing server logic using a web server and communicating via HTTP)
- Adding file handling (e.g. reading/writing spreadsheets or CSV files).
- storing and querying data from a database (nontrivial database structure with a number of connected entities and queries with multiple joins)
- generating reports or charts from data,
- implementing some non-trivial algorithms or data structures such as Dijkstra's algorithm, or a physics simulator, or writing your own linked list/tree
- Adding a user login and user profile system with the ability for users to register themselves
- User roles (e.g. administrator, manager, user, guest) which lead to different permissions for different features of the app.
- Security features such as password complexity, retries/lockout, password reset, storing encrypted data, etc.
- Working with some kind of web API, or maybe sending e-mails from the app.
If you'd like to do something which could lead towards AI, then a chatbot or a game could be a good place to start. There are a lot of online guides and tutorials about writing simple chatbots so you could look into those. If you ended getting a simple chatbot working, then maybe you could extend it by using some internet-based AI tools like Text-To-Speech or even Speech Recognition (again, you can find guides and tutorials for using these; maybe using Google's AI tools)
Alternatively, you could look at a card game or something like battleships -- it's generally easier to start with a game which has no AI at all and simply allows players to play each other, then build some kind of simple AI into it if you still have time and you need to add more complexity.
Whatever you do, make sure that you focus plenty of effort into writing the report and on your objectives and success criteria before you start writing out any code; the initial phase of writing the report makes a big difference to being able to write the code. The time you spend analysing and decomposing the main features and capabilities of the app will make a big difference when it comes to writing the code - i.e. being able to have a bunch of bullet-pointed lists which describe what the program is actually expected to do, how users will interact with it, describing/drawing the UI/UX, and then thinking really carefully about how you are going to test and verify each of your objectives.
Doing this kind of thinking before you write any code should give you a clear vision about what you're actually creating, and it's all really important for the report too, which is a really significant part of the overall project, and is more likely to make the biggest difference to your final score for the coursework. (making sure you read the mark scheme carefully and in particular not dropping any easy/obvious marks for the report)