Regex is a little quicker than the accepted answer (for my 23 MB test file) that I used. But there isn't a lot in it.
import re
bad_words = ['bad', 'naughty']
regex = f"^.*(:{'|'.join(bad_words)}).*\n"
subst = ""
with open('oldfile.txt') as oldfile:
lines = oldfile.read()
result = re.sub(regex, subst, lines, re.MULTILINE)
with open('newfile.txt', 'w') as newfile:
newfile.write(result)