Whether you can use AI in an internship application depends on the employer and the application stage. Check the policy before using it to draft, edit or practise. A company can value AI skills at work while requiring you to complete its recruitment assessments independently.
UK internship and graduate job applications. Employer guidance checked on 23 September 2026.
Can you use AI for internship and graduate applications?
There is no single rule across employers. Oxford's careers guidance recommends checking each organisation's instructions, retaining your own examples and verifying information generated by a tool. It also distinguishes job applications from further-study applications, which have their own policies. Oxford's guidance on AI in applications is a useful starting point.
For a concrete employer example, PwC's UK early-careers assessment policy requires independent work. Unless PwC explicitly authorises it, AI tools or third-party assistance must not generate, edit, rewrite, summarise, coach or otherwise complete assessment responses. That restriction is broader than asking a tool to write the final answer. Read PwC's assessment and selection policy and the instructions supplied for your stage.
Do not assume another employer uses the same rule, or that a rule about assessments settles what is permitted during preparation. If the wording does not answer your situation, ask the recruitment team before using the tool.
Separate the stages before deciding
| Stage | What to establish first |
|---|---|
| Research | Can you use AI to explore a sector, and which facts need checking against official sources? |
| Written application | Are drafting, grammar edits or other assistance permitted? Is disclosure required? |
| Practice | Which preparation materials and types of coaching does the employer allow? |
| Live assessment | What tools, notes and assistance are explicitly permitted during this task? |
| Interview | What are the rules for preparation, notes and assistance during the conversation? |
Save the policy link and the date you checked it in your application checklist and tracker. Recheck the instructions when an assessment invitation arrives: it may contain requirements that were not on the original vacancy page.
Keep an application rooted in your actual work
Start with your own evidence before opening a writing tool. Note the situation, what you personally did, the result and what you learned. Our student skills evidence worksheet gives you a structure for that first draft.
Where assistance is permitted, compare every suggested revision with the original facts. Watch for small changes that inflate your role:
- “Helped organise” becomes “led” even though someone else managed the event.
- “Compared two designs” becomes “optimised performance” without a test result.
- “Built a coursework prototype” becomes “delivered a production system”.
Put the accurate wording back. A stronger verb is only useful when it describes what happened. Keep estimates, measured outcomes and personal reflections distinct.
For employer research, follow a generated suggestion to its original source. Check the organisation, programme, office and intake year. A plausible description of a closed programme will still send your application in the wrong direction.
How to explain an AI-assisted project
Using AI in a project and using it in an assessment are separate questions. For a project, first follow the rules of the course, employer or organisation that owns the work. Then explain both the assistance and your own contribution.
Here is an illustrative independent project: a student investigates library opening hours using public information. They use a coding assistant to suggest an initial spreadsheet-cleaning script, then review and test it themselves.
| Project note | What the student could record |
|---|---|
| Task | Compare published opening hours across a defined set of libraries |
| Tool contribution | Suggested a first version of the code used to read the spreadsheet |
| My decisions | Chose the comparison period and kept holiday exceptions separate |
| My checks | Tested missing values and checked several records against the source pages |
| What changed | Corrected code that treated a blank closing time as midnight |
| Limits | The comparison describes published hours, not how frequently people use each library |
The useful detail is the judgement: noticing an ambiguous value, checking the source and changing the approach. If you cannot explain a line of code or a conclusion, investigate it before claiming it as evidence of your skill.
For an interview, practise describing one decision, one error you found and one limitation without the tool. Follow the employer's rules for that preparation and the interview itself.
Before you submit
Use this short review alongside the employer's instructions:
- Have I followed the policy for this exact stage?
- Can I substantiate every claim about my role, skills and results?
- Have I checked employer facts against the current official page?
- Have I kept confidential material and other people's personal details out of tools and public examples?
- Have I included any disclosure the employer requires?
- Could I explain this work in my own words?
If you already have project material to share, our internship portfolio guide explains how to select relevant examples. Compare FolioBuild's portfolio templates for a layout that makes the task, contribution and outcome easy to find. Whatever tool helps you present the work, the final account should remain accurate and recognisably yours.