The lexer sees an f-string:
f'some text {something.split('
next to a normal string:
')}'

The first of those is not a valid f-string because of the unmatched left brace, so it's an error.

I'm contemplating making the f-string parser smarter to be able to detect this, but it's a long way off.

For an explanation of why it works the way it does currently, see https://mail.python.org/archives/list/python-dev@python.org/message/BDZCXGRW5KTUOGMRT6OHH6S3UD4BV5ZV/ . This post also has an example similar to yours.