Not an expert on this, but infants show "racial" bias to the race of their parents, not their own race. Studies with adopted infants of different races from the parents have shown this. This is thought to be a visual familiarity effect babies associate between caregivers and those who look similar to caregivers, which makes sense as others that are close to the infant are likely to be extended family of similar ethnicity.

Even so, this doesn't prove babies are inherently racist, only that they associate visual information from those around them to possibly categorize who is family and who is not. It should also be noted that reasoning for familiarity vs novely effects in infants is a bit murky, and as far as I know researchers have a hard time determining why some studies return familiarity effects and why some show novelty effects (where infants are more interested in something new)

He then makes this grand leap to saying this is related to cops shooting different races at different rates. Again, not an expert, but doesn't the data show that white cops are less likely to shoot at black perpetrators than black cops are to shoot at black perps? He's taking an idea that isn't even correct, linking it with another idea that isn't correct, and then saying "this is the only reasonable explanation" and that it's all CRT

This was longer than I wanted to write out, but I wanted to make it clear how wrong @ryu289 is here once again.