As a non-native English speaker, I am often a little bit unsure about grammar, wording, or phrasing of my posts. Ignoring the fact that this might be hard to detect by others: if I ask an LLM like ChatGPT to improve my post in a linguistic-only fashion, without changing the meaning, does the result then count as "AI generated content"?
I think we should allow these use cases.
FYI, for reference, some research venues explicitly allow these use cases. E.g. ACL 2023 (which is a top-tiered conference for natural language processing): https://2023.aclweb.org/blog/ACL-2023-policy/: (authors: ACL 2023 Program Chairs)
Here is our take on some cases frequently discussed in social media recently:
Assistance purely with the language of the paper. When generative models are used for paraphrasing or polishing the author’s original content, rather than for suggesting new content - they are similar to tools like Grammarly, spell checkers, dictionary and synonym tools, which have all been perfectly acceptable for years. If the authors are not sufficiently fluent to notice when the generated output does not match their intended ideas, using such tools without further checking could yield worse results than simpler-but-more-accurate English. The use of tools that only assist with language, like Grammarly or spell checkers, does not need to be disclosed.
Short-form input assistance. Even though predictive keyboards or tools like smart compose in google docs are also powered by generative language models, nobody objected to them, since hardly anyone would try to use them to generate a long, unique and coherent text: it would simply not be practical. Similarly to language tools above, the use of such tools does not need to be disclosed in response to the writing assistance question.
Literature search. Generative text models may be used as search assistants, e.g. to identify relevant literature. However, we expect the authors to read and discuss such references, just like the references identified by a regular search engine or a semantic literature recommendation tool. The usual requirements for citation accuracy and thoroughness of literature reviews apply; beware of the possible biases in suggested citations.
Low-novelty text. Some authors may feel that describing widely known concepts is a waste of their time and can be automated. They should specify where such text was used, and convince the reviewers that the generation was checked to be accurate and is accompanied by relevant and appropriate citations (e.g., using block quotes for verbatim copying). If the generation copies text verbatim from existing work, the authors need to acknowledge all relevant citations: both the source of the text used and the source of the idea(s).
New ideas. If the model outputs read to the authors as new research ideas, that would deserve co-authorship or acknowledgement from a human colleague, and that the authors then developed themselves (e.g. topics to discuss, framing of the problem) - we suggest acknowledging the use of the model, and checking for known sources for any such ideas to acknowledge them as well. Most likely, they came from other people’s work.
New ideas + new text: a contributor of both ideas and their execution seems to us like the definition of a co-author, which the models cannot be. While the norms around the use of generative AI in research are being established, we would discourage such use in ACL submissions. If you choose to go down this road, you are welcome to make the case to the reviewers that this should be allowed, and that the new content is in fact correct, coherent, original and does not have missing citations. Note that, as our colleagues at ICML point out, currently it is not even clear who should take the credit for the generated text: the developers of the model, the authors of the training data, or the user who generated it.
A separate, but related issue is use of generative models for writing code. ACL submissions may be accompanied by code, which counts as supplementary materials that the reviewers are not obliged to check and consider, but they may do so if they wish. The use of code assistants such as Copilot is also a relatively new practice, and the norms around that are not fully established. For now, we ask the authors to acknowledge the use of such systems and the scope thereof, e.g. in the README files accompanying the code attachments or repositories. We also ask the authors to check for potential plagiarism. Note that the Copilot in particular is currently the subject of a piracy lawsuit, and may have suggested snippets of code with licenses incompatible with yours. The use of code assistance does not obviate the requirements of authors to ensure the correctness of their methods and results.
And CEUR-WS proceedings. https://ceur-ws.org/ACADEMIC-ETHICS.html:
In the past few months, we have witnessed the emergence of novel large language models (LLM) reaching breakthrough performance on NLP tasks. These include ChatGPT and Galactica, which are AI assistants that can produce long and good quality content that can be seeded for authors’ work. Because of their recent emergence, the norms around the use of such technology is not fully established, yet. Hence, it is important to acknowledge its use and elaborate on how it has been employed.
Specifically, we define three levels of AI assistance usage: insignificant, low and substantial. We will group the different use cases according to these three categories and we will define CEUR-WS stance.
Insignificant. Activities like: i) paraphrasing and refining the manuscript content (using Grammarly or other spell checkers), and ii) smart composition (via predictive keyboards) are widely accepted and do not need any acknowledgement.
Low. The use of AI tools for searching and generating literature review is acceptable upon authors’ checks. Authors must review the content and adjust/add references to line up with the narrative of their manuscript. In case of generating unoriginal content (i.e., definition, or description of well-known concepts) may be acceptable provided that the authors have checked it to be accurate and included proper references to the original content.
Substantial. Using AI assistants for generating new ideas as well as new text is unacceptable. Most of the generated content may derive from existing work. Potential issues with such practice are related to originality, plagiarism, ownership, and authorship, whose consequences and impact are not yet clear.
Regardless of the cases above, CEUR-WS publishes original work from named authors, and thus contributions from AI assistants can only be stated in the acknowledgements and/or by suitable references at the original research papers. We require that all authors and workshop editors adhere to these guidelines. Their violation will lead to the removal of the published paper or the whole volume, similar to our procedures dealing with plagiarism.
As this technology is in current development, we plan to continuously review this policy in the upcoming months.
This policy section is partly inspired by the “ACL 2023 Policy on AI Writing Assistance” available here.
Related documents:
- US Copyright Office's Guidance on AI-Generated Material (2023-03-16)
And I'd like to add a note on Tinkeringbell's answer:
I'd ask you (and everyone else who's considering doing this) why you think there's a need to use the LLM for this. It seems like a case of 'when you have a hammer, everything looks like a nail'. LLMs like ChatGPT aren't specifically designed as tools for spelling/grammar checking. It just takes a prompt and uses some kind of statistical relationships it learned during training to make some text.
Not being designed to be a specific tool doesn't make them bad. E.g., LLMs are not designed to do text summarization specifically, however they are state-of-the-art text summarizers. Therefore, the argument is invalid.
