I want to rename S3 objects atomically.

It's gzip compressed in a specific bucket, but the extension is saved as .json.
If things go on like this, the batch process will not recognize it as a gz file, so
I want to rename .json.gz

However, it is already in service and there is always a write from the API in this bucket
json.gz presence check → renaming is not atomic, so

if update processing from API is interrupted during rename processing after existence check I will overwrite the latest data with old data, so I would like to prevent that from happening

If it were a file system, I would do something like lock.
Are there any locks on S3 objects?

Or, it seems that the renaming user can create but not update
I thought it would be possible if I had IAM permission.
S3:PutObject permission appears to be the only one for object writing

Could you give me some advice on what to do in this case?