Google drops pledge not to use AI for weapons or surveillance

Is this more or less ethical than OpenAI getting a DoD contract to deploy models on the battlefield less than a year after saying that would never happen, with the excuse being well we only meant certain kinds of warfare or military purposes, obviously. I guess my question is, isn't there something more honest about an open heel-turn, like Google has made, compared to one where you maintain the fiction that you're still trying to do the right thing?

I think it's unfair to bring up OpenAI's commitment to its own principles as any sort of bar of success for anyone else. That's a bit like saying "Yes, this does look like they're yielding to foreign tyrants, but is this more or less ethical than Vidkun Quisling's tenure as head of Norway?"