Microsoft has become the latest company to have ended up in the list of Big Tech firms that listen in on their users from various sources and transcripts. According to a report by Vice, Microsoft has hired contractors through a third party entity, which listen in on bits of collected Skype conversations, and even snippets of commands recorded by Cortana.
The motive, as other firms have also claimed, is to reduce various aberrations such as false keyword activation, consistency of voice-driven services, accuracy of responses given and so on. However, in order to properly tabulate these bits of information, tech firms also include identifiers in these recordings, which despite not directly identifying a particular user, recognises aspects such as time, place and location of recording, which still puts sensitive private information at risk.
The risk arises from the third party human involvement, which is particularly risky since many of these third party human service providers are not properly authorised, and might hence end up misusing the power of information available to them. Many experts have stated that instead of human involvement, companies should look at implementable AI models for this task, which would also help them maintain the data in-house instead of outsourcing them.
Microsoft, it is important to note, states clearly in its Skype privacy documentation that some bits and pieces of conversation data may be recorded and used to improve the service — something that Google and Amazon clearly faced a setback on. Given the increasing amount of scrutiny here, it remains to be seen how this segment of voice data usage evolves, in due time.