I think that what you're seeing here is a conference chair who doesn't understand what arXiv is and is blindly applying the self-plagiarism policy.

The IEEE's current official FAQ on author rights has a question dedicated to arXiv that explicitly states that arXiv publication is permitted. Thus, there is no concern on that account.

The IEEE CrossCheck system, however, does not filter arXiv out of its web-crawling. Thus, if you've posted something to arXiv, CrossCheck will dutifully report that it has found a high similarity match to text found on arXiv. And this is where I think that things have gone wrong.

• What CrossCheck actually does is report materials that should be examined to see whether they are (self-)plagiarism or not.
• People who use CrossCheck sloppily, however, often don't bother with examining, and just assume that match = plagiarism.

Thus, I think your colleague is dealing with somebody who hasn't bothered to understand what CrossCheck found and is just reaching for a knee-jerk reject. I notice as well that the email you quote doesn't say how, exactly, the rules are being violated.

From this comes my recommendation for how to proceed. I would recommend writing back to the conference chair to ask for clarification, while pointing at the IEEE FAQ, something like:

My apologies, but I am confused. The IEEE author rights FAQ explicitly allows for posting of preprints on arXiv, so I don't think that can be the violation of IEEE rules that you are referring to. I don't think this can be a self-plagiarism issue since arXiv isn't a peer-reviewed publication. Does this conference have a different policy, and if so, can you please point me to which aspect my preprint is violating?

Hopefully, this will either lead the conference chair to understand that they have made a mistake or else point you to the actual issue that they have. If they're a martinet or a fool, however, they may still just demand the thing gets taken down, in which case you can attempt to do the arXiv withdrawal process, making sure to point the blame at the conference in the comments field.

Update based on response from conference chair:

The uninformative and inflexible response from the conference chair does not affect my advice: I believe it simply means you are going down the "martinet or fool" path. Get the withdrawal done, making sure the comments point squarely at inflexible orders from the conference chair. Get the publication completed and officially published on IEEE Xplore. Then, if possible, put the preprint back up on arXiv where it belongs and see if your colleague can put a word in the ear of IEEE or senior colleagues who can appropriately castigate the conference chair for being a fool and an ass.

I think that what you're seeing here is a conference chair who doesn't understand what arXiv is and is blindly applying the self-plagiarism policy.

The IEEE's current official FAQ on author rights has a question dedicated to arXiv that explicitly states that arXiv publication is permitted. Thus, there is no concern on that account.

The IEEE CrossCheck system, however, does not filter arXiv out of its web-crawling. Thus, if you've posted something to arXiv, CrossCheck will dutifully report that it has found a high similarity match to text found on arXiv. And this is where I think that things have gone wrong.

• What CrossCheck actually does is report materials that should be examined to see whether they are (self-)plagiarism or not.
• People who use CrossCheck sloppily, however, often don't bother with examining, and just assume that match = plagiarism.

Thus, I think your colleague is dealing with somebody who hasn't bothered to understand what CrossCheck found and is just reaching for a knee-jerk reject. I notice as well that the email you quote doesn't say how, exactly, the rules are being violated.

From this comes my recommendation for how to proceed. I would recommend writing back to the conference chair to ask for clarification, while pointing at the IEEE FAQ, something like:

My apologies, but I am confused. The IEEE author rights FAQ explicitly allows for posting of preprints on arXiv, so I don't think that can be the violation of IEEE rules that you are referring to. I don't think this can be a self-plagiarism issue since arXiv isn't a peer-reviewed publication. Does this conference have a different policy, and if so, can you please point me to which aspect my preprint is violating?

Hopefully, this will either lead the conference chair to understand that they have made a mistake or else point you to the actual issue that they have. If they're a martinet or a fool, however, they may still just demand the thing gets taken down, in which case you can attempt to do the arXiv withdrawal process, making sure to point the blame at the conference in the comments field.

Update based on responses from conference chair and arXiv:

The uninformative and inflexible response from the conference chair does not affect my advice: I believe it simply means you are going down the "martinet or fool" path.

Since arXiv has (appropriately) refused to take the publication down, there is now no path left but confrontation with the program chair. IEEE might be able to help in time, but you're dealing with a large and often slow-moving organization there, so if there isn't a response from IEEE within a day or so, it's going to have to be dealing with the conference instead.

Here, it may be useful to involve other people besides just the program chair. Most well-established conferences have some sort of steering committee or similar that is specifically designed to help ensure consistent behavior despite the year-to-year change in organizers. Writing to them to ask for help with dealing with the failure of the specific chair that your colleague is dealing with may be useful.

I think that what you're seeing here is a conference chair who doesn't understand what arXiv is and is blindly applying the self-plagiarism policy.

The IEEE's current official FAQ on author rights has a question dedicated to arXiv that explicitly states that arXiv publication is permitted. Thus, there is no concern on that account.

The IEEE CrossCheck system, however, does not filter arXiv out of its web-crawling. Thus, if you've posted something to arXiv, CrossCheck will dutifully report that it has found a high similarity match to text found on arXiv. And this is where I think that things have gone wrong.

• What CrossCheck actually does is report material that should be examined to see whether they are (self-)plagiarism or not.
• People who use CrossCheck sloppily, however, often don't bother with examining, and just assume that match = plagiarism.

Thus, I think your colleague is dealing with somebody who hasn't bothered to understand what CrossCheck found and is just reaching for a knee-jerk reject. I notice as well that the email you quote doesn't say how, exactly, the rules are being violated.

From this comes my recommendation for how to proceed. I would recommend writing back to the conference chair to ask for clarification, while pointing at the IEEE FAQ, something like:

My apologies, but I am confused. The IEEE author rights FAQ explicitly allows for posting of preprints on arXiv, so I don't think that can be the violation of IEEE rules that you are referring to. I don't think this can be a self-plagiarism issue since arXiv isn't a peer-reviewed publication. Does this conference have a different policy, and if so, can you please point me to which aspect my preprint is violating?

Hopefully, this will either lead the conference chair to understand that they have made a mistake or else point you to the actual issue that they have. If they're a martinet or a fool, however, they may still just demand the thing gets taken down, in which case you can attempt to do the arXiv withdrawal process, making sure to point the blame at the conference in the comments field.

I think that what you're seeing here is a conference chair who doesn't understand what arXiv is and is blindly applying the self-plagiarism policy.

The IEEE's current official FAQ on author rights has a question dedicated to arXiv that explicitly states that arXiv publication is permitted. Thus, there is no concern on that account.

The IEEE CrossCheck system, however, does not filter arXiv out of its web-crawling. Thus, if you've posted something to arXiv, CrossCheck will dutifully report that it has found a high similarity match to text found on arXiv. And this is where I think that things have gone wrong.

• What CrossCheck actually does is report materials that should be examined to see whether they are (self-)plagiarism or not.
• People who use CrossCheck sloppily, however, often don't bother with examining, and just assume that match = plagiarism.

Thus, I think your colleague is dealing with somebody who hasn't bothered to understand what CrossCheck found and is just reaching for a knee-jerk reject. I notice as well that the email you quote doesn't say how, exactly, the rules are being violated.

From this comes my recommendation for how to proceed. I would recommend writing back to the conference chair to ask for clarification, while pointing at the IEEE FAQ, something like:

My apologies, but I am confused. The IEEE author rights FAQ explicitly allows for posting of preprints on arXiv, so I don't think that can be the violation of IEEE rules that you are referring to. I don't think this can be a self-plagiarism issue since arXiv isn't a peer-reviewed publication. Does this conference have a different policy, and if so, can you please point me to which aspect my preprint is violating?

Hopefully, this will either lead the conference chair to understand that they have made a mistake or else point you to the actual issue that they have. If they're a martinet or a fool, however, they may still just demand the thing gets taken down, in which case you can attempt to do the arXiv withdrawal process, making sure to point the blame at the conference in the comments field.

Update based on response from conference chair:

The uninformative and inflexible response from the conference chair does not affect my advice: I believe it simply means you are going down the "martinet or fool" path. Get the withdrawal done, making sure the comments point squarely at inflexible orders from the conference chair. Get the publication completed and officially published on IEEE Xplore. Then, if possible, put the preprint back up on arXiv where it belongs and see if your colleague can put a word in the ear of IEEE or senior colleagues who can appropriately castigate the conference chair for being a fool and an ass.