The second of our principles states: Robots should be designed and operated to comply with existing law, including privacy. Yes, sometimes the obvious does have to be stated.
We also worried about hacking. Our third principle is: Robots are products: as with other products, they should be designed to be safe and secure. And the commentary for this principle has this to say about hacking:
We are aware that the public knows that software and computers can be “hacked” by outsiders, and processes need to be developed to show that robots are secure as far as possible from such attacks.It seems our concerns were well founded. Last week a report appeared of a networked baby monitor that was apparently hacked. It was pretty distressing. The hacker was shouting abuse at the baby, chillingly using her name - it seems that he (let's assume it was a he) was able to gain access to the baby monitor's video feed and read the baby's name displayed above her bed. Even more chilling (adding to the parent's horror) the execrable hacker then turned the camera to look at them when they entered their child's room to find out what was going on.
WiFi IP Camera, aka Baby Monitor is, I contend, a teleoperated robot. The thing that makes it a robot is the motorised pan and tilt mechanism for steering the camera. So, despite password protection, Mr Gilbert's networked robot was hacked. The hack was a clear violation of both privacy and security. This particular robot, and I hazard hundreds of thousands like it, is absolutely not secure from attack. It fails our 2nd and 3rd principles.
The consequences of this particular attack were, fortunately, not much more serious than giving the Gilbert's a fright they surely won't forget for some time. But for me one particularly egregious aspect of this robot hack - something that the robot ethics working group did not anticipate - was the verbal abuse hurled at baby Allyson and her parents. It is with profound dismay that I ask the question: is the first case of RoboTrolling..?