UnTXT and Imperial College London's AI Policy - What Students Need to Know

Summary: Imperial's policy requires explicit authorization before using AI in assessed work. Unauthorised use may be treated as contract cheating. UnTXT helps you identify and remove AI patterns from drafts, ensuring your final submission meets Imperial's standards for substantial personal contribution.

What Imperial's AI Policy Actually Says

Imperial's answer to "can I use this?" is, more than at most universities, your department's answer.

Its own checklist for students leads with exactly that instruction: "Review your department's current policy on using and disclosing genAI in academic work." Imperial's college-level principles set direction and say plainly that they "are not intended to be exhaustive", noting that the use of generative AI "will vary across academic disciplines." Aeronautics and Medicine are not obliged to land in the same place, and they do not.

That devolution is the single most useful thing to know as an Imperial student, and it is the thing most third-party summaries of Imperial's position leave out.

What this means for a detection tool. UnTXT does not generate anything — it reads a draft you already wrote and marks passages carrying machine-writing patterns. Most departmental policies are written about tools that produce text, which is a different object. But we are not going to tell you your department has approved it, because we have not read your department's policy and Imperial's own guidance is that it is the one that governs.

Find it, and read the paragraph about disclosure specifically.

What we would do. Say you used it. One line — "I used UnTXT, an AI detection tool, to check my own draft before submission" — costs nothing and settles it. Imperial's general expectation is that AI-derived material in your work is attributed; a line about a tool you only used to check yourself sits comfortably inside that habit.

Source: Imperial Generative AI Checklist and Generative AI Principles, checked 7 September 2026. Your department's policy sits above both. Where Imperial's pages and this one disagree, theirs is the one that counts.

Understanding "Explicit Authorization"

Imperial's requirement for explicit authorization means:

Check these sources (in order):

1. Your assessment brief - Does it explicitly permit AI use? Some briefs say "you may use generative AI for brainstorming only" or "AI is not permitted." Read carefully.

1. Your module guide - Does your module leader specify AI policy? They may have given guidance at the start of term.

1. Your department - Email your module leader or department if the assessment brief is unclear. Get written confirmation.

1. Imperial's general guidance - The university's policy provides a framework, but individual modules may be more or less restrictive.

If you can't find explicit authorization, assume it's not permitted. Better to ask than to guess and risk academic misconduct findings.

Why Imperial Students Use UnTXT

You used AI for brainstorming—but you rewrote it, and you need to prove it

Even if your module doesn't allow AI use in the final submission, many allow it for brainstorming. If you used AI as notes and rewrote substantially in your own words, UnTXT shows you exactly which passages still have AI characteristics. This gives you confidence that your revision is thorough.

You got explicit authorization for limited AI use—but you want to stay within bounds

Your module leader approved AI for outlining or first-draft generation, but the final submission must be substantially yours. UnTXT helps you identify any passages that aren't substantially revised, so you can fix them before submission.

You're writing under time pressure and want to ensure you don't accidentally cross a line

When deadlines are tight, it's easy to paste AI output or lightly edit it without thinking. Running UnTXT before submission catches these moments and gives you a chance to revise properly.

How to Use UnTXT for an Imperial Assessment

Before you start: Confirm the AI policy for your specific assessment. Don't assume.

If AI is not authorized:

1. Write your draft entirely yourself (or use AI for notes only, then rewrite)
2. Read the grey passages - these are the ones still carrying machine patterning after your rewrite
3. Re-run UnTXT to confirm resolution
4. Submit with confidence - Your work is substantially your own

If AI is authorized (for specific purposes only):

1. Use AI only for the authorized purpose (brainstorming, outlining, etc.)
2. Rewrite fully in your own words - This is non-negotiable
3. Run UnTXT to identify any passages that still carry AI patterning
4. Revise those passages before submission
5. Acknowledge the authorized use in your statement: "I used AI for brainstorming, then rewrote entirely in my own words. I used UnTXT to audit the final draft for AI patterns."

What the Results Mean for Your Imperial Submission

UnTXT's flags indicate passages with AI writing characteristics. At Imperial, where authorization is required, any substantial AI-patterned text in your final submission is problematic—unless you have explicit permission and your use is limited to what was authorized.

The modules most relevant to Imperial assessments:

A cluster of grey passages is worth a second read. It does not establish that anything was left unrewritten — formal academic prose produces these patterns on its own, and plenty of entirely human paragraphs come back grey. It tells you where to look, and you are the one who knows what happened when you wrote it.

Frequently Asked Questions

What counts as "explicit authorization"?

Something in writing that clearly permits AI use for that specific assessment. Email confirmation from your module leader is explicit. The assessment brief saying "you may use AI for brainstorming" is explicit. Assuming it's okay because a friend did it is not explicit.

What if my module guide doesn't mention AI at all?

Email your module leader and ask. Most will clarify. If they don't respond, assume it's not permitted.

If I used AI without authorization, what should I do?

First: Don't submit it as-is. Second: Rewrite the passages substantially in your own words. Third: Run UnTXT to confirm they're revised. Fourth: If you're past that point, consider contacting your department to disclose before they discover it. Most institutions treat disclosure far better than discovery.

What if UnTXT flags my essay but I didn't use AI?

It happens. Formal academic writing, hedging language, and quoted material can trigger flags. That's why you review the flagged passages in context. If they're genuinely your work, they should be defensible. If you're unsure, revise them to match your established voice more clearly.

Can I explain to Imperial that I used a humaniser tool?

Not advisably. Using an evasion tool after unauthorised AI use makes the integrity violation worse, not better. Better to rewrite authentically and acknowledge what happened if asked.

How much does UnTXT cost?

First 1,000 words free. Most essays fit within that. For regular or higher-volume use, individual accounts move onto a low-cost monthly subscription (from £8.40/month, tax-inclusive) with a monthly CVT allowance.

If Your Work Is Questioned

If Imperial's academic integrity team reviews your work:

1. Be completely honest - Tell them exactly what you did. Hiding makes it worse.
2. Provide evidence - Earlier drafts, notes, timeline of creation
3. Run UnTXT yourself ahead of time - Show them you audited your own work
4. Demonstrate revision - If you had authorization for brainstorming, show how you rewrote substantially
5. Reference Imperial's policy - You were operating within (or complying with) the framework

Most cases resolve through conversation if you're honest and can demonstrate good faith effort to comply with the policy.

Why Authorization Matters

Imperial's requirement for explicit authorization isn't bureaucratic gatekeeping. It reflects the principle that your assessed work should demonstrate your understanding, not an AI's output.

When you're authorized to use AI for brainstorming but must rewrite substantially, you're learning the material. When you're not authorized and you use it anyway, you're not learning—you're just outsourcing.

The integrity standard protects not just the institution's reputation, but your own education.

The Bottom Line

Imperial's policy is clear: unauthorized AI use is serious. But the policy also recognises that AI is a tool, and there are legitimate ways to use it within appropriate bounds.

UnTXT helps you stay within bounds. By auditing your work before submission, you ensure it meets Imperial's standard for substantial personal contribution.

Start for free - first 1,000 words. No payment needed. Sign up at untxt.uk

This page reflects UnTXT's understanding of Imperial College London's publicly available AI policy as of early 2026. Policy details may change. Always check Imperial's official guidance and consult your department or module leader for your specific assessment.