IMO nothing “has” free will objectively. It’s observer-dependent, in that “Does this have free will?” is exactly equivalent to asking “Can I perfectly predict the behavior of this?”