[D] Advice for spotting "fake" ML roles?

I recently got hired for what turned out to be a fake ML role even by the loosest definition of an ML role (Bloomberg AI group). It seems like currently there are many companies/teams/people who pretend to do ML work and when it comes to hiring they lie to candidates about the work they actually do. Does anyone have any strategies to spot these kinds of roles so they can be avoided? It seems like asking questions during the interview is not very effective because you could easily just be lied to.