For a student or researcher considering the use of Generative AI applications in their work, it is important to know the difference between Gen AI-Assisted and Gen AI Generated. Students may be allowed to use Gen AI for assistance, but not use it to Generate work.
For example, Amazon now requests publishers submitting their work to declare whether it is AI 'generated' or 'assisted' in its new policy for Kindle authors.
The important overall difference between the two is the level of human involvement, with more human involvement in the ultimate output where it is only Gen AI assisted and less involvement where the output is entirely Gen AI generated.
Gen AI - Generated
We define Gen AI-generated content as text, images, or translations created by an AI-based tool. If you used an AI-based tool to create the actual content (whether text, images, or translations), it is considered “AI-generated,” even if you applied substantial edits afterwards.
If a Gen AI tool is the primary creator of content, then it becomes the author, and it is Gen AI generated.
Examples would be:
Gen AI Application to write a book review from scratch, or to answer an essay question.
Asking Gen AI to create a picture of a red robot sitting on the moon.
Gen AI - Assisted
If you, as the author, created the content yourself, however used Gen AI based applications to edit, refine, error-check, or otherwise improve that content (whether images or text), then this would be considered “Gen AI assisted” rather than “Gen AI generated.” You could also use Gen AI based tool to generate new ideas or brainstorm concepts, but if you ultimately created the text or images yourself, this would be considered “Gen AI assisted” and not “Gen AI generated.” It is not necessary to inform us of the use of such tools or processes, but you should always reference them.
If a Gen AI tool is used to improve, filter, modify or enhance the work a student or user creates, then that
would be regarded as Gen AI Assisted.
An example would be asking a Gen AI application to change a photo to black and white, to check spelling in a document, or
change the verb tense from passive to active in your writing.
This table shows comparative examples:
| Gen AI Assisted | Gen AI Generated | |
| Topic Selection and refinement | Students use prompts with suggestions and provides criteria to Gen AI which then suggests potential topics or refines existing topics | AI selects a topic without suggested topics or input |
| Images | Request a Gen AI application to change a colour photo to black and white | Ask Gen AI to create a picture of a red robot sitting on the moon |
| Planning and Outlining | Gen AI application organises a student's notes that they have prepared into an outline. | AI organises notes not written by the student (eg tutor or lecturer supplied notes) into an outline |
| Manuscript/Book writing | Gen AI provides suggestions for variations on a theme that the writer has created. | AI writes the entire manuscript from scratch. |
| Editing/Revising a work | Generative AI suggests changes to improve pacing or structural edits. | AI is asked to re-write a book to improve pacing or structure. |
| Editing/Proofreading | Gen AI suggests changes to switch from passive verbs to active verbs. | AI translates a student's text from their native language into English. |
This table is adapted from the AI Assisted vs AI generated module table in Epigeum's/Clarivates Academic Integrity module.