2024-2025 AMCAS Certification Update - Use of AI

[WildWing]

An eagle-eyed pre-health advisor noticed that there was an update to the AMCAS certification statement in regards to AI:

I certify that all my writing, including personal comments, essays for MD-PhD applicants, and descriptions of work/activities, is my own. Although I may utilize mentors, peers, advisors, and/or AI tools for brainstorming, proofreading, or editing, my final submission is a true reflection of my own work and represents my experiences. I acknowledge that no changes can be made after submission and will thoroughly proofread my work. Quotations are allowed if I cite the source.

For those folks who have expressed concerns about using AI to help with development of their personal statements and essay responses, this should put your mind at ease as long as you didn't put "write me a personal statement for medical school" into ChatGPT and then cut and pasted directly into AMCAS.

---

Members don't see this ad.

 

[pablo213]

An eagle-eyed pre-health advisor noticed that there was an update to the AMCAS certification statement in regards to AI:

I certify that all my writing, including personal comments, essays for MD-PhD applicants, and descriptions of work/activities, is my own. Although I may utilize mentors, peers, advisors, and/or AI tools for brainstorming, proofreading, or editing, my final submission is a true reflection of my own work and represents my experiences. I acknowledge that no changes can be made after submission and will thoroughly proofread my work. Quotations are allowed if I cite the source.

For those folks who have expressed concerns about using AI to help with development of their personal statements and essay responses, this should put your mind at ease as long as you didn't put "write me a personal statement for medical school" into ChatGPT and then cut and pasted directly into AMCAS.

This is a pretty vague certification statement. What exactly is the line between "editing" and something being "a true reflection" of your work? If I'm prompt engineering to get a certain output, is that not a reflection of my work and thought process in some capacity? If I "edit" every single sentence in my essay, is it still my own work?

Either way, this still seems to be an honor based agreement that can't really be upheld one way or another, since there's no way to definitively prove AI use. On another note, I do recall seeing individual schools with policies regarding AI use for secondaries. I wonder how suspected AI-use will affect applicants next cycle (or if it already has this cycle).

thoughts @eigen ?

 

[Mr.Smile12]

This is a pretty vague certification statement. What exactly is the line between "editing" and something being "a true reflection" of your work? If I'm prompt engineering to get a certain output, is that not a reflection of my work and thought process in some capacity? If I "edit" every single sentence in my essay, is it still my own work?

Either way, this still seems to be an honor based agreement that can't really be upheld one way or another, since there's no way to definitively prove AI use. On another note, I do recall seeing individual schools with policies regarding AI use for secondaries. I wonder how suspected AI-use will affect applicants next cycle (or if it already has this cycle).

thoughts @eigen ?

Not going to argue. The Code of Conduct is an honor code/agreement, just as most professional codes of conduct are. You won't get arrested for violating it, but you will likely be prohibited from the only gateway to being a physician (or a professional for other applications).

I have said that all the application essays should be well-written and edited following feedback from references and others. Was there ever any prosecution of candidates who may have "plagiarized" other applicant essays? No one has the time.

I think what is worthwhile is that this statement opens a door for dialogue about how admissions committees will monitor AI usage. We know from TurnItIn data over the years the detectors are not definitive (see the article I posted about TurnItIn's analysis of papers it received over the last year about AI usage). I will say that long-term, this will help schools lean into SJTs more. What will be more interesting is how AI could help admissions offices score your application... but that's a conversation for next year or so. (I have already remarked how AI can be useful to calibrate screeners reading your secondary essays.)

 

[RunnerDad]

Not going to argue. The Code of Conduct is an honor code/agreement, just as most professional codes of conduct are. You won't get arrested for violating it, but you will likely be prohibited from the only gateway to being a physician (or a professional for other applications).

I have said that all the application essays should be well-written and edited following feedback from references and others. Was there ever any prosecution of candidates who may have "plagiarized" other applicant essays? No one has the time.

I think what is worthwhile is that this statement opens a door for dialogue about how admissions committees will monitor AI usage. We know from TurnItIn data over the years the detectors are not definitive (see the article I posted about TurnItIn's analysis of papers it received over the last year about AI usage). I will say that long-term, this will help schools lean into SJTs more. What will be more interesting is how AI could help admissions offices score your application... but that's a conversation for next year or so. (I have already remarked how AI can be useful to calibrate screeners reading your secondary essays.)

At the higher ed district where I provide legal counsel, we are struggling with the same issues of clarity around expectations for students. I think AAMC's decision to focus on the applicant's certification that their submission is a "true reflection of [their] own work and represents [their] experiences" is the right call, because it gets away from a line-by-line technical analysis of whether specific words and phrases were adopted from other sources (whether human or generative AI). Even a document written entirely by AI consists of words and phrases that are in common usage (especially if the prompt writer doesn't say, "Explain it like I'm a Ph.D."), and it's very likely that I might use phrases and sentences that have been recycled over decades or centuries, such as with metaphors. This is why faculty at my institution have been instructed that analyses by software like TurnItIn cannot be used as a basis in and of itself to discipline students for academic misconduct.

Where things get really sticky is when a policy leaves too much discretion to the faculty member to interpret vague or overbroad definitions, such as when a policy says a student may not submit something that they didn't "substantially complete." What does that even mean? Believe it or not, there are several dozen committee members, stakeholders, lawyers, and administrators battling on that front where I work, and it's a mess. Students must be on fair notice of what they can and cannot do, and anytime a policy is left to the broad interpretation of faculty or administrators, there is a real danger that some will use the policy as a weapon against students they just don't like -- whether it's because the student is a jerk (understandable, but still not a good reason for an academic misconduct IA), or they have a certain skin color or gender (which is illegal and even more reprehensible).

The dialogue is good, and AAMC has at least recognized that the use of AI is not only inevitable but can be a positive thing in improving clear communication.