The Apple HomePod And Siri: A Match Made In Heaven... Or Hell

Apple's new HomePod sees the company joining the home concierge business, expanding the footprint of Siri from traditional computing devices into a home appliance. But is that a good thing? Is Siri really reliable enough for your home.

In Apple's press release, that hit my inbox this morning, the company said

Security and privacy are fundamental to the design of Apple hardware, software and services. With HomePod, only after “Hey Siri” is recognised locally on the device will any information be sent to Apple servers, encrypted and sent using an anonymous Siri identifier.

On face value, that all sounds reasonable. Siri won't collect any of your requests and pump them up to Apple's big virtual brain unless you invoke Apple's version of "abracadabra" and say "Hey Siri". But here's the thing, all of my Siri-enabled products react to phrases other than "Hey Siri". I can be driving and chatting and my Apple Watch will buzz, telling me Siri couldn't recognise what I was saying. Other times, my iPhone does the same.

If we can't trust how the devices react to random audio that may or may not sound anything like Hey Siri, then how can we stop Apple from collecting random bits of audio. Even data that it can't use is personal data that is being collected without permission because the permission-seeking tool, "Hey Siri" malfunctions.

I might be paranoid but I'm not sure the world really needs the ilk of Amazon Alexa, Google Home or Apple HomePod listening in to everything just in case it hears the right magic word. And, while their intentions may be noble, I'm not particularly trusting of these devices yet. Although there haven't been any major breaches yet, I'm certain there are bad guys trying to exploit them and there are hackable flaws that haven't yet been revealed.

There have been cases of microphone-enabled devices capturing data and not handling it securely, such as last year's CloudPets fiasco. I'm confident that Apple and the others are doing things better than the CloudPets mob but that doesn't mean they are infallible.

Who's got one of these home assistants? Do you trust them? Or are you limiting what they can do somehow?


    If someone hacks Amazon, Google or Apple I think there's bigger issues to worry about than random voice samples from the general public...

    I’m glad I’ll get to hear “sorry, I can’t help you with that” im high def.

    The privacy concerns need breaking into two categories. The first is the idea of a device listening all the time (and only uploading to the web after a trigger is heard). Those concerns can be addressed by the tech-savvy owners checking their internet routers and seeing what the "connected home" devices upload and when. Any irregularities would likely gain the warranted press attention.
    The second is what happens to the data that's uploaded. As in the case of the referenced CloudPets case, that's all to do with how uploaded data is held by the entity it's uploaded to. Do we trust Google/Amazon/Apple to behave ethically, legally, responsibly and correctly with whatever voice snippets are uploaded to them? I've not read all their legal/privacy agreements etc, but I would imagine their stance is probably pretty much the same as for conducting the same search/command via their more usual non-voice recognising services.

    I would trust Apple long before I would trust Google or Amazon.

    Hell, Google and Amazon can’t be trusted with your data.
    That’s a simple fact based on their business model - you are the product to be sold to advertisers.
    They will never respect your privacy.

Join the discussion!

Trending Stories Right Now