The ethical questions raised by Google's (small) contract with the US military to develop AI are difficult and controversial, but they have to be asked, and i'd much rather they're asked in a form open to public debate than behind closed doors.
https://www.theverge.com/2018/5/30/17408446/google-ai-guidelines-weaponry-military-pentagon-maven-contract
Where will Google draw the line on weaponized AI?