Academic Honesty and AI

 Integrity is the backbone of science. As a scientific publisher, eiRxiv therefore follows strong ethical guidelines. These ethical guidelines help us ensure the manuscripts we publish are accurate, truthful, ethically done, and gives credit where needed. 

We encourage you to thoroughly review our guidelines on Plagiarism, Academic Honesty, Conflicts of Interest, and AI use below. 

Not following these guidelines may result in your manuscript failing pre-review or being publicly withdrawn. Authors who seriously violate these academic honesty expectations may be banned from all future publishing opportunities with eiRxiv and JEI. 

Make sure your work is entirely your own original creation and give credit when using someone else’s ideas, results, or images. Basically, if it’s not common knowledge or your own original idea, you need to cite it. eiRxiv does not tolerate plagiarism, which is using another person’s work without crediting the source. 

That said, we also understand that plagiarism can sometimes happen unintentionally – not because authors are intending to plagiarize, but because they simply don’t understand what does or doesn’t constitute plagiarism and don’t take enough steps to prevent it. 

At eiRxiv and JEI, manuscripts that have not properly cited their sources will be sent back to the authors and–in severe cases–may be rejected, whether or not the plagiarism is intentional. 

Review JEI’s Academic Honesty page for more information on what plagiarism is, how to avoid it, and how to incorporate and cite your sources.  eiRxiv holds manuscripts to the same academic honesty standards as JEI. 

 eiRxiv also does not tolerate authors making up data or results, or falsifying any supporting documents (signatures, agreements, or permissions). 

Falsifying results seriously damages the trust and reputation of scientists as individuals and the scientific process as a whole and can have legal consequences for the authors. Maintaining scientific integrity is extremely important to be respected and valued within and beyond the scientific community. 

You should also be honest about your hypothesis. Make your hypothesis before you start your experiments or analysis, and be honest about whether or not your results support your hypothesis. 

Do not change your hypothesis after you get your results. It’s ok to have an unsupported hypothesis, but it’s not ok to be dishonest about it. 

  A conflict of interest (COI) occurs when an author or someone they’re affiliated with financially benefits (or seems to benefit) from a manuscript being published. 

Examples of COIs include but are not limited to: 

• A soda company publishes a study showing that in a blind taste test, participants preferred their soda over a competitor’s soda. 
• Student authors are testing the health benefits of a tea produced by the company that one of the adult authors works for. 
• Researchers are being paid by an aeronautics company to test a new airfoil model. 
• The senior author’s spouse is on the board of directors for a company that develops location-based tracking software. The student author used data from this company’s software in their research. 
• One of the adult authors on a study showing airbags decrease head injuries during car crashes was paid to be a pro-airbag expert in a lawsuit against a car manufacturer. 

Conflicts of interest matter because science is meant to be unbiased. If a COI exists, the author may consciously or unconsciously change their experiments, results, analysis, or interpretation to benefit themself or someone they know. In addition, readers need to know if a COI exists so they can evaluate whether the research is trustworthy. 

If a COI is discovered after a manuscript has been published, the manuscript may be retracted. Retracting articles due to undeclared COIs can damage public trust in science. 

At eiRxiv, we require authors to declare all Conflicts of Interest. 

• If one of your authors has a COI to declare, please complete our Conflict of Interest form below and submit it with your manuscript. 
• If you have a COI and don’t declare it, this is considered a breach of publication ethics and may result in disciplinary action. 
• If you’re not sure whether or not you have a COI to declare, please reach out to us at submissions@eirxiv.org.
• Whether or not you have a COI, you will need to complete the “Conflict of Interest” section on the manuscript template. Please see Write Your Manuscript for more information on how to format this COI declaration. 

While we understand why AI can be a helpful tool in some instances, our mission is to support students as they develop their own scientific writing skills and unique scientific voice. We believe first-hand experience with developing ideas, creating hypotheses, conducting research, and writing about it is the best way for students to develop as scientists and communicators. 

In addition, AI tools sometimes pull text, images, and other information from other published, copyrighted, or private work without properly citing it, notifying the AI user, or getting permission from the original creator. 

It is also important to remember that AI and other similar tools (Large Language Models, Machine Learning, etc.) generate results based on predictions. Because of this, these models are known to “hallucinate,” or make up results that simply sound or look good even if they are incorrect. 

As scientific publishers, we have an ethical responsibility to publish only research that is scientifically accurate and does not infringe on anyone else’s copyright. We therefore cannot ethically publish manuscripts that use AI in the following ways: 

• Reading, summarizing, and analyzing previously published scientific studies
• Writing text or making revisions (exception: AI tools can be used for proofreading minor spelling and grammatical mistakes after you write your text. However, they shouldn’t be used while you write). 
• Creating images
• Conducting statistical analysis, modeling, or calculations 
• Data mining (searching for patterns in large datasets without a hypothesis about what you will find)
• Citing references or creating reference lists

In short, your work must be your own original creation. You must read and summarize previous studies, make your hypothesis, perform the experiments, analyze and interpret data, draw conclusions, generate images, and cite resources yourself. 

For additional information on manuscripts that use AI as the tool for conducting experiments, visit Topics We Accept (link).