There are enormouse people in the world of who's English is not their first language. They have some problem in writing a paper with correct English. But if the sense is ok and merit is upto the mark then why paper acceptence gets a question?
Because 'errors' in written English can be very distracting to the reader and impede appraisal of the paper. For papers written in English that is not fluid and error-free but otherwise acceptable, I usually recommend 'minor revisions' and suggest that the author have a fluent English writer or translation service help the writer revise the paper. I often take a paragraph from the paper & cast it into more professional English to demonstrate the types of changes that are needed. It might be possible to use an AI programme as a translation service, although I have no experience in that.
Now English check is more easier but it is related with the financial issues. You know it is difficult for a early researcher to manage this issue so why not the publisher take the responsibitilies of it.
Rajat Sanker Roy Biswas Many universities that offer remedial English classes or English for Nonnative speakers will help with translations or even improving the written English of native English speakers who are faculty members. I have frequently referred PhD students to the remedial English writing services even when spoken English is not a problem for them. I agree that it is expensive to engage professional translation services, but perhaps some AI function can rewrite passages for free. I hesitate to assign the responsibility to publishers when they are already charging exorbitant fees for open access publication and for selling journals to academic libraries.