What are the odds that those robots have perfect software security and every individual who ever worked on them will not leak source/specs/backdoors/admin passw

What are the odds that those robots have perfect software security and every individual who ever worked on them will not leak source/specs/backdoors/admin passwords once they see them used oppressively? Just like they aren't human and don't feel empathy, I don't feel empathy for shutting down a robocop or subverting his programming to ignore me or sending it on a suicide mission back to its makers.