10

Is it safe to write research paper on Overleaf? Some of my friends say that it is not safe to write research papers on it as it is connected it internet. Please suggest. I am afraid of plagiarism.

12
  • 27
    Plagiarism? This is quite different from someone stealing your username/password and stealing your work. You do know that even writing your papers locally has the same risk? Nowadays with malwares, once you digitize your thought, it can be stolen. To push it further, any nearby cameras can be hacked to take a snapshot of your paper. I doubt anybody would want to steal your papers unless your topics are sensitive or have commercial value. In both cases, you wouldn't use a cloud-based system. Commented Mar 25, 2021 at 6:38
  • 2
    @RobbieGoodwin Standards differ across disciplines, remember? In many areas of CS, LaTeX is fast becoming a defacto standard. Many journals and conferences provide their own standardized templates following their publication requirements. I know of academics that refuse to review manuscripts which are not typeset in LaTeX. And while I personally think that is extreme, I certainly know that I shan't be submitting again to the journal that forced me to submit in MS Word (specifically, forced me to switch in the middle of the review process), and most certainly not as corresponding author.
    – penelope
    Commented Mar 26, 2021 at 14:07
  • 12
    Hello, Tom from Overleaf Support here. Note that while Overleaf is indeed operated over the internet, we do take data privacy and security very seriously. You can read more about it at overleaf.com/legal#Privacy and we are happy to answer your questions at [email protected] . (Note that I don't make a full answer here as I only deal with one part of the whole consieration, and the existing answers tackle the other parts very nicely.)
    – yo'
    Commented Mar 26, 2021 at 14:07
  • 10
    @penelope: "LaTeX is fast becoming a defacto standard..."? Maybe if you wrote this sometime in the last century, but LaTeX has been the de facto standard since before there even WAS a MS Word.
    – jamesqf
    Commented Mar 26, 2021 at 16:23
  • 3
    The typical procedure of collaborators e-mailing Word files back and forth is way easier to hack than Overleaf.
    – Joooeey
    Commented Mar 28, 2021 at 16:21

3 Answers 3

62

Overleaf is unlikely to be a problem

To plagiarize your work, someone would have to

  • find it very valuable
  • hack your Overleaf account (and know it is there they need to hack) or that of your coauthors or you would have to have shared it publicly
  • finish the research before you do, which typically means they would have to have superior or in any case solid knowledge in the field,
  • publish it before you do
  • have more credibility than you do in the academic community, so that you can't just say that it was actually your work they are plagiarizing

Of all these factors the second one is the only one that has anything to do with Overleaf. In order for it to be the crucial issue, you already need to have a competent and well-connected researcher out to plagiarize you. While it could happen, it feels unlikely, and you probably know if you are in this situation or not.

If you do happen to be in a position that requires extreme information security (governmental research, dual purpose research, authoritarian regime, a powerful enemy out to get you), then it is a good rule of thumb to not use Overleaf or other cloud services. But in this case you really should seek help that is a lot more competent and specific than the answers here.

6
  • The risk with a plagiarist finishing and publishing first could be more severe if the OP is trying to perfect a paper (perhaps by acquiring one last result) and submit to a slow, high-impact journal. The plagiarist could then go for a quicker journal, one that's barely respectable, happy to publish the more preliminary results, but included in the major indices. It's still highly unlikely though
    – Chris H
    Commented Mar 26, 2021 at 11:06
  • 6
    "hack your Overleaf account" -- this depends on how the Overleaf project has been set up. Using free accounts, often the only way to work collaboratively is to allow anybody with a link to the project to access the text. The security is then provided by the long-and-random nature of the project's URL, but if the link is intercepted (say, a coauthor inadvertently forwarding an email with the link to someone they shouldn't have) then the text could leak, potentially to other groups in a better position to plagiarize.
    – E.P.
    Commented Mar 26, 2021 at 13:31
  • "finish the research before you do, which typically means they would have to have superior or in any case solid knowledge in the field," But if the part that the author on Overleaf had done was the hard part, requiring ingenuity in addition to solid knowledge, then the plagiarist might need only solid, but far from superior, knowledge. Commented Mar 27, 2021 at 18:13
  • @MichaelHardy Right; superior knowledge, or at least solid. If my language is easy to misunderstand, feel free to clarify with an edit.
    – Tommi
    Commented Mar 28, 2021 at 7:58
  • 2
    @Tommi Don't you think Neelkanth was asking not all about being plagiarised, but "I am afraid of being accused of plagiarism"? Commented Mar 28, 2021 at 20:03
25

A wide range of Academicians and Researchers have been using overleaf for a long time now, none of them have complained of anything as such other than some suggestions of improving it. I've personally never had any issues with overleaf since it is much more convenient to have a common platform for collaboration. Also, being insecure about an idea/paper being stolen isn't something one should be bothered about unless you've a ground breaking idea (pun unintended) However, if one is reluctant to use the cloud based LaTeX editor, one can always switch to an offline LaTeX software if one wants. That should help.

4
  • 10
    Actual problems experienced with Overleaf include: 1. unexpected downtimes before a deadline; 2. weird behavior if some collaborators use their git integration and some don't Commented Mar 25, 2021 at 16:50
  • 3
    I do agree with that. Also, I remember a couple of researchers last month who were actually working in parallel and at the same time, complained of losing progress of what they had actually written and that too abruptly. Not once, but multiple times! Commented Mar 25, 2021 at 17:01
  • @AaqibBashir merging simultaneous changes is theoretically doable (manually) but it's generally best to try and avoid working on the same section at the same time, and an author needs to take on the duty of integrating changes.
    – Chris H
    Commented Mar 26, 2021 at 11:07
  • @ChrisH Thanks for these insights! Quite valuable indeed! Commented Mar 26, 2021 at 12:00
3

Overleaf is indeed slightly more insecure than Dropbox, Git or e-mail, because it has one additional avenue for attack: you can get illicit access to a document stored there if you obtain the 24-digit hex number that appears in its URL. Clearly the probabilities to guess one at random are minuscule, but there are in theory some exploits that could help (sniff one from a plain-http connection, visited link attacks, etc.).

Dropbox also has unauthenticated share tokens, but typically one shares their document with specific users anyway.

Of course this is all theoretical, and probably this risk is only a tiny fraction of that coming from more traditional attacks (weak passwords and phishing, for instance --- as the saying goes, "problem in chair, not in computer").

2
  • 1
    You can restrict access by overleaf user instead of relying on a secret link. I don't think any of my overleaf documents have link sharing on.
    – user133933
    Commented Mar 30, 2021 at 6:44
  • I believe you need a paid plan for that, though, and most researchers don't have that. Commented Mar 30, 2021 at 6:48

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .