Looks like a healthy plant.
Stick with it, but sometimes autos aren't autos, more often from bad breeding techniques .. it all comes down to math and a resulting %. it's not quite 100% no matter how sophisticated the process, i'd assume. if i knew how many alleles were invovled and what not, you could do the factoring in punnett squares and determine after X generations what the % would be that retains the autoflowering capacity.
Anyway, give it time.. may be a nice big autoflower.. who knows? But at some point you'd want to treat it like a photoperiod...