The Oxbridge Editing Blog 26th June 2024

How to Proofread and Edit AI-Generated Content

26th June 2024
Speak right now to our live team of English staff

While writing your academic work, you might have generated some parts by AI. This is understandable since AI produces content quickly and can save you lots of effort and time. However, you are well aware that AI is permitted by universities only when it is used correctly. Generating work with AI is prohibited and penalised. For this reason, you will want to proofread and edit AI-generated content as a means of reducing the risk of plagiarism and ensuring that your work aligns with academic integrity standards. Here we provide some tips that can help you in the process.

Proofread to Reduce Language Complexity

If some parts of your work were generated by AI, you can be sure they are free of spelling, punctuation, and grammatical errors. This is because AI tools, such as ChatGPT, write flawlessly. While proofreading AI-generated content, your goal will not be to correct such errors but to identify and reduce all instances of language complexity. The reason for this is that ChatGPT uses complex language that could be easily picked up by your supervisors, who could then flag your work as AI-generated. During the proofreading process, be sure to locate complex words such as “ubiquitous”, “exacerbate”, “leverage”, and “nefarious”. In fact, be sure to simplify the language wherever you feel that such simplification would ensure that your work reads naturally.

Proofread to Ensure Alignment with your Writing Style

Paragraphs or sections produced by AI could easily stand out if they do not align with your writing style, as evident in paragraphs or sections written by you. While proofreading, you will want to identify all such inconsistencies and correct them. For instance, if you notice that AI-generated parts lack the typical tone or vocabulary you use, revise them. Similarly, if you see that these parts have a tone or vocabulary you do not usually use, be sure to rewrite them. During the final proofread, make sure that all parts of your academic work read as if they were written by you. 

Edit to Reduce the Bland and Generic

Despite their flawless language skills, AI tools such as ChatGPT produce bland and generic content. Be aware of this as such content is a direct giveaway of AI overuse. Examples are expertly written sentences that do not signal specificity and critical thinking nor consider context and examples. Basically, these are perfectly written sentences that do not say anything of value or advance the argument. Be sure to identify all such AI-generated sentences or paragraphs and enrich them. Populate them with specific examples, personal insights, or well-considered perspectives that demonstrate deeper understanding and engagement with the topic.

Edit to Reduce the AI Score

Your supervisor may decide to run your work through an AI detector to see how much of it was human-written. Truth be told, these detectors are highly unreliable as they often flag student-written work as AI-generated. However, to be on the safe side, you may decide to run the parts written by AI through an online detector, such as ZeroGPT. If these parts return with a high AI score, you will want to edit them to reduce the score. The more you edit these parts and add your personal touch to them, the less will they get flagged as AI-generated.

Takeaway Message

Overusing AI while writing could lead to academic penalties. However, if you did use AI to help yourself out, you can avoid penalties and potential academic integrity loss by proofreading and editing AI-generated content. Some tips are to proofread to reduce language complexity and ensure alignment with your writing style, as well as to edit to reduce bland and generic content and lower the AI score. What you should avoid at all times is submitting AI-generated work without sufficient proofreading and editing as this could lead you to get caught.