Master of Laws (LL.M.) Statement: Who Can Help

Can You Get Help With a Master of Laws (LL.M.) Personal Statement?

Sometimes — and at four of the best-known Master of Laws (LL.M.) destinations, no. Harvard, Columbia, Duke, and Cambridge each publish a rule that restricts outside help on this exact document, and Harvard and Duke both make you certify authorship when you submit. Other programmes publish nothing on the subject, which is not the same as permission.

So the question is never "is getting help allowed in general." It is: what does my programme's own sentence prohibit, and does what I am about to do fall inside it? Here are the sentences, checked on August 16, 2026.

What the programmes actually publish

If you are applying to Harvard, Columbia, Duke, or Cambridge, do not run that draft through an outside review of any kind, including ours. That is not hedging. Harvard's sentence reaches "any other person," Columbia's reaches "assistance in completing any components," and both reach AI by name.

The certification is the part people miss

The published policy is only half the instrument. The other half is what you personally attest to at submission.

Harvard's required certification reads: "I certify that the information presented in my application is accurate, complete, honestly presented, and the writing is entirely my own and not that produced by any other person or artificial intelligence program."

Duke's certification form is signed the same way: "I certify that all information submitted as part of my application is true and complete to the best of my knowledge, and that my personal statement and all other writings are my own work, drafted by me and without assistance from artificial intelligence technology."

Note what those change. A policy on a webpage is the school's statement. A certification is your statement, made under your own name, and it travels with the file. If you have already used something that a certification would make untrue, the problem is not detection — it is that you would be signing something you know to be false.

Read the certification inside your own application before you submit it, not the summary of it on the admissions page. And where a general policy and a programme instruction disagree, follow the stricter one.

Duke draws the line most usefully

Duke is the one programme in this set that bans AI assistance and, in the same breath, says what human help remains available: its rule "does not preclude asking family members, friends, and others for proofreading assistance or general feedback."

That is the distinction the whole question turns on, written by a school with one of the stricter rules:

• Feedback and proofreading on writing you produced — contemplated.
• Drafting — "you wrote all drafts and the final product."

The same shape appears at Columbia, which explains its ban in terms of what the document is for: "You should prepare this document without the assistance of others, including AI technology, because this will not provide us with the personal information we seek." The rule is not there to catch you. It is there because a statement that is not yours does not tell them the thing they are reading it to learn.

Cambridge goes further than the others in one specific direction, and foreign-trained applicants should read it carefully: the prohibition "covers the use of AI to help your use of English." Language polishing is not carved out. If you are a non-native English speaker applying to Cambridge, the ordinary workaround is not available to you there.

Critique versus ghostwriting

Once you are outside the programmes that ban assistance outright, the meaningful line is not "human versus AI." It is who produced the words on the page.

Everything on the left returns you to your own draft with work to do. Everything on the right hands you sentences you did not write — and the right-hand column is what every rule quoted above is aimed at, whether the words it uses are "written by," "assistance," or "author."

Worth noticing while you shop: a service that promises to produce the statement for you is offering the right-hand column regardless of whether a person or a model does the producing. Harvard's and Columbia's rules name both. If an offer's core promise is that you end up with text you did not write, the programmes above have already answered it.

Silence is not permission

We searched the current LL.M. and graduate requirement pages for Georgetown, UCL, King's College London, and LSE on August 16, 2026 for "artificial intelligence," "generative AI," and "ChatGPT," and found no AI rule for this document on any of them. Do not read that as a green light, for three reasons:

1. The own-work requirement can exist without the word "AI." LSE's supporting documents page requires that statements are "all your own work. If we discover this is not the case, your application may be cancelled" — and, in the same list, that they are "proofread before being uploaded." Both obligations are live whether or not a policy names a technology.
2. The certification inside the portal can be stricter than the public page. We could not read every school's in-application certification from outside its portal, and that is exactly where authorship promises usually live.
3. Rules change between cycles. Every school in this article can revise its text before your deadline. The page that governs you is the one live when you submit, not the one quoted here.

The safe default when a page is silent: treat the document as work that must be demonstrably yours, and if the stakes make you uncertain, email the graduate admissions office and work from what they send you. A written answer from the office beats any inference, including ours.

A five-minute check before you accept any help

1. Open your programme's own application-instructions page, not the university's general integrity policy.
2. Search it for "AI," "artificial intelligence," "generative," "assistance," "consultant," and "own work."
3. Identify the verb in whatever you find: written by, assistance in completing, used when writing, author. "Assistance" is the widest — a tool that only comments on finished writing is still assistance you sought.
4. Open the certification you will have to sign and read it in full.
5. If either instrument reaches what you were about to do, the answer is no. If they conflict, follow the stricter one.

The same method works across the rest of the law application; we walk through it verb by verb in can you use AI on a law school essay, and the attestation side is covered in what you are actually signing in an AI attestation.

Where a rubric review fits, and where it does not

GradPilot reads writing you already wrote. It does not draft, does not rewrite, and does not return replacement prose to paste in. What comes back from the Master of Laws (LL.M.) personal statement rubric is criterion-based critique tied to your own sentences, plus authenticity signals describing how the text reads — the same posture as the rest of the law rubrics.

Where your programme permits review of your own draft, that is what it is for. Where your programme does not — Harvard, Columbia, Duke, and Cambridge, on their current published rules — the answer is no, and no feedback is worth an integrity finding.

Three limits belong in the same breath. Scores are revision signals, not admissions predictions. An AI-sounding flag is a reason to inspect a passage, not proof of how anything was written. And no review certifies compliance with any programme's rule — that determination is yours, made from your programme's current words and the certification you sign.

Related reading

• Master of Laws (LL.M.) personal statement guide — the four jobs and the published limits
• The LL.M. personal statement for foreign-trained lawyers — the licensing route and what it changes
• Can you use AI on a law school essay? — reading a policy sentence for its verb
• What you are actually signing in an AI attestation
• AI law school personal statement review: how it works — what a rubric review does and does not do

Sources

All retrieved August 16, 2026. Assistance and AI rules change between cycles; confirm your programme's current wording and read the certification inside your own application before you submit.

• Harvard Law School 2026–2027 LL.M. application instructions (PDF) — the own-efforts rule and the required certification
• Columbia Law School LL.M. application process — the strict prohibition and the reason given for it
• Duke Law LL.M. application information — the generative-AI rule and the proofreading carve-out
• Duke Law international graduate programs certification form (PDF) — the signed authorship certification
• University of Cambridge — using AI in your application — the personal statement prohibition and its English-language clause
• NYU School of Law LL.M. required materials — the authorship rule
• NYU School of Law graduate admissions ethical conduct — third-party consultants "strongly discouraged"
• LSE supporting documents — own-work and proofreading requirements
• Georgetown Law LL.M. programme-specific requirements, UCL Laws LL.M. applying and entry requirements, and King's College London Master of Laws entry requirements — checked for AI and assistance rules; none found on these pages on the date above