Each time you use a video teleconferencing app, you send your audio data to the company hosting the services. And, according to a new study, that means all of your audio data. This includes voice and background noise, whether you are broadcast or mute.
Researchers at the University of Wisconsin-Madison examined “many popular apps” to determine the extent to which video conferencing apps capture data while users use the in-software ‘mute’ button.
According to a university press release, their findings were substantial:
They used runtime binary analysis tools to detect raw audio in popular video conferencing applications as the audio traveled from the app to the computer audio driver and then to the network while the app was muted.
They found that all the apps they tested occasionally collected raw audio data while mute was enabled, with one popular app collecting information and delivering data to their server at the same rate, regardless of whether the microphone was muted or not.
Unfortunately, because this study remains unpublished, we are unable to confirm the specific test results. That, for now, we can not call them and shame.
However, the effectiveness of this paper is not necessarily in doubt due to the fact that it has been accepted for the 2022 Privacy Enhancing Technologies Symposium. We’ll just have to wait and see who’s name-dropped when the newspaper comes out in June.
However, that does not mean that we can not draw some conclusions. According to the researchers, this data can be used to extract meaningful information. And, with a little machine learning, that information can paint an incredibly vivid picture of a user’s reality – again, even with your microphone silently in the app.
The research team was able to determine which specific audio was sent during testing and, by extrapolating this data, they were able to predict what conclusions large tech could make.
Of course, big tech uses AI to parse everything. That the researchers built their own algorithms to study the data. What they found was staggering.
According to the abstract of the unpublished paper:
Using network traffic that we intercept on the way to the telemetry server, we implement a proof-of-concept background activity classification and demonstrate the usefulness of deriving the ongoing background activity during a meeting – cooking, cleaning, typing, etc. We achieved 81.9% macro accuracy in identifying six common background activities with intercepted outbound telemetry packets when a user is muted.
In other words: graduate students at the University of Wisconsin-Madison were able to build machine learning models that were able to determine what a teleconferencing app user was doing while their microphone was muted with more than 80% accuracy.
File under: Imagine what Google or Microsoft could do with their massive AI models. Yikes!
Neural syn take: Do you need to worry? Yes. Absolutely. Do you need to stop using these apps? No, because that’s not really an option for everyone.
Our real concern is that great technology does not matter whether users know they are being recorded, even when they are silent, or that it never stopped to think that users soe care. However, it shows a disturbing level of detachment from the user experience.
Big tech’s unspoken mantra is “data for any price,” and this is just to prove how thirsty the beast is. There is no good reason not to explicitly inform users in large print that the mute button does not stop their audio feed to the server.
Fortunately, you have options. If you really want to mute your audio feed, you need to perform a double mute. If you are lucky enough to own a headset that has a physical mute button on it, use it to mute your headset and then use the button in the app as an extra layer of muting.
If your headset does not have a physical mute button, or you use a microphone on board to communicate, you need to mute the system by mute the system settings in your operating system, and mute in the operating system. app.
At the end of the day, it’s unclear exactly what big tech is doing with this data. The scope of the study did not include any major tech research – and we’ll update this piece if Zoom, Microsoft, or Google return with a statement.
But, regardless of what, the fact that it is collected under such misleading circumstances are cause for great concern.
Forcing users to go into operating system menus to ensure they have a little privacy is an anti-user policy. And moreover, it shows how much more sensitive our audio data can be than our video data.
As lead author of the study, Kassem Fawaz said in the university’s press release:
With a camera you can turn it off or even put your hand over it, and whatever you do, no one can see you. I do not think that exists for microphones.
Do not forget to double mute, people. You will probably forget to double down switch on from time to time, but the variation keeps Google, Microsoft and everyone else from training their machines on the environmental noises of your private life.