Helping a politician who ends up killing people is arguably helping to kill those people, whether you're a voter or an ad firm or IBM in the 40s.
What should the standard of proof be for accusing an organization of “helping to kill”? Have we met that with Facebook?