As automation becomes more advanced, less expensive, and easier to obtain, more and more duties will be given to machines that were once the job of live humans. There are indeed many benefits to taking people out of harm’s way or off of menial tasks, and it is a process unlikely to slow down.
Though a robotic security officer may be able to patrol and report some types of incidents completely autonomously, that is only effective if it doesn’t decide to drive itself into a pool of water as the model in Washington DC recently did. Whether that should prove to be the baptism at the birth of AI security, or proof that robots can never fully replace humans, it shows that there is still a long way to go before the constructive reasoning and decision making of the human brain can be removed from the equation.
Full reliance on machines can become an issue with systems much less complicated than synthetic patrol officers, which has also been highlighted in recent events, including the Australian traffic camera system that was infected with a virus, and sent out thousands of false tickets automatically. Even with no Internet connection, a maintenance mistake is believed to have affected hundreds of cameras. These are but two of many examples to illustrate that while technology is improving, it cannot be relied on without human oversight as a failsafe, and that if an incident or intrusion does occur, there is no substitute for having a trained responder on site at the property.