Victim of article theft wants correction to list their name, not retraction

Case number:

Case text (Anonymised)

Author A contacted us claiming that an article published in the journal recently by author B was stolen from an article author A had earlier submitted to two different publishers, publisher A in 2016 and publisher B in 2017. Author A provided the PDFs of the manuscripts they had submitted to those other publishers. The version submitted to us 2018 by author B was very similar to that submitted to publisher B.

We contacted publisher B who confirmed the details of the submission to them by author A in 2017. Author B is listed publicly as a reviewer for publisher B's journal, but publisher B could not confirm that they had direct access to this particular submission. Author B said their PhD advisor, now apparently deceased, had given them the article but they recently had doubts that this had been their advisor's work. They agreed to retraction.

Author A has asked whether instead of retracting we might publish a correction to replace author B with the rightful author, Author A, because the article has already been peer reviewed and accepted.

Question(s) for the COPE Forum
• Is correcting a stolen article to list the correct authors a potential solution? If so, should we correct the article metadata too?
• Is there any precedent for such a total correction of authorship?
• How might we detect and prevent the publication of stolen articles? They do not show up on Crossref Similarity Check because they are unpublished.