I think some people may not trust talking to robots with internet access like how some people do not trust using a internet connected security cameras, and speakers like Amazon Alexa which send the users text, audio and video recordings to a internet company like Amazon who can share user's recordings with the government, advertisers, or another company like Twitch and Facebook.
But, there are most likely a lot of people who may already trust robots enough to tell their secrets, and embarrassing stories to a robot since a lot of people already use smartphones and computers with voice assistants like Apple Siri, Google Now, and Microsoft Cortana to search for info on almost any topic.
I feel it is not a good idea to tell robots secrets because the robot's operating system maybe infected with spyware which is listening to people's conversation through the mic, and secretly making video recordings through the back and front camera, and sending the recordings to a company or the government.
I think when robots look more like humans, many people may accidentally tell a robot their secret because they thought the robot was their human friend, or a traditional human worker like a therapist where you sometimes need to tell them your secrets, so they can help you deal with your mental health problems, and the human-like robot would leak human's private conversations if the robot got hacked, has spyware, or the government and companies has direct access to the robot's computer hardware and software through an internet connection.
There may also be a risk that a human talk to a robot bird, dog and cat, and told the robot animal a secret which get leaked to a hacker or sent to a big company or the government.
You shouldn't trust any device with a microphone or camera. If it can see or hear you then assume that it is always watching/listening.
And as more devices are capable of connecting to the internet (or require an internet connection to function) even less trust in them should be placed. I mean can you say with any certainty that the device you own/are using isn't watching you right now? Listening to every word you say? And would you even be able to tell if someone was actually tuning in? (The answer to that one is: not likely.)
Consider things like Google Home or Alexa... they are always listening whenever plugged in so they can hear their specific keywords to respond. So even in cases when you may not want them to hear what you're saying they're still listening. Can the owner also hear in this mode? Supposedly no... but there's no way to tell what is and isn't getting transmitted back to the parent company in some form. (What's sent back might not be voice clips but may just be text transcripts of things you've said. They might be kinda garbled, but if the machine can accurately respond to what you're asking for and give an answer it can probably write down something close to what you asked for.) So even if not instructed to listen it might still be there silently gathering data on you. So if you talk with a friend about looking at buying a new vehicle you may find ads for vehicles show up more frequently for you because the device heard you say you were looking or thinking about it.
And that's just for rather mundane stuff. But what about more personal info? Things like medical records/history (what prescriptions you're on, diseases or illnesses you have, etc...) or financial records. And what if you've done something wrong (even accidentally) like say reveal a company secret? Or worse something actually illegal?
Look at your cellphone? I mean they have cameras/microphones and many have GPS locators on top of the wi-fi, 3g or 4g data and the regular cell service. So not only can your device track you pretty easily via GPS, but it can do so via cell tower triangulation (which are accurate within roughly 5 meters). This means any cellphone can hear you, see you, track you and report everything back to a third party.
So be careful what you say and to whom. Big Brother is watching and listening. George Orwell's 1984 isn't so far fetched now with the technology we have.
I feel many robots may not use a secure operating system, or the maker of the robot's OS would eventually stop making security updates for the robot's operating system, so it can be more vulnerable to hackers and viruses like Wannacry which can automatically infect vulnerable systems to install ransomware and other unwanted software like spyware/malware.
"$26 (£16) software let militants view potential targets
• American official says flaw was identified and fixed
One of America's most sophisticated weapons in the conflicts in Iraq, Afghanistan and Pakistan, the unmanned drone, has been successfully penetrated by insurgents using software available on the internet for $26 (£16)."
I don't have high expectations that average robots will have good security since Drones which cost more can be hacked with a $26 hacking software tool.
