Forum Moderators: open
If you're running A/B tests, then it shouldn't cause a problem. I assume the redirect you've already put in place redirects people to a new page. You simple make that new page the A/B test. I'm mostly used to running multivarient tests with optimiser, but I suspect you'd be OK with A/B as well.
One solution would be to run your A/B as a multivarient. So, rather than 2 separate pages and worring about redirects just run 2 big variations on the same page. That way all should be fine.