In both cases you aren't really sure it's the case, you wouldn't know if it were suddenly disregarded or circumvented, and you're just trusting in the good behavior of your chosen provider.
However, we all make certain risk tradeoffs based on numerous things, including proxy signals. For instance, I would be intensely surprised to find out my downstairs neighbor was recoding me, based on multiple years of interactions. Likewise, I find the signal Apple sends ("We need a US-based GDPR") about privacy rather more compelling than the signal Amazon sends ("scour the mic for hints about product we can shift").
I mean, sure, while both $huge_bank and your hypothetical deadbeat relative might both steal your money, I bet I know which one you choose to to leave your money with.
Totally tangential, so feel free to disregard:
This statement reminded me of my experience during my brother's TS clearance (Canada). CSIS readily had accurate data (in surprising detail) about my time spent during a short period they must have observed his family members to use during an interview with him. It was pretty fascinating, mildly disconcerting— but in this case didn't bother me (in the context, I understood why. I certainly wasn't interesting at the time). They only could have captured the data they had by a combination of tailing (however brief or otherwise more expensive much more technical means) and bank records. I was of course none the wiser.
Advertisers are another matter. I have a Google Home Mini. It remains unplugged 100% of the time we're not using it. My phone can always go in another room. If we decide to turn it on to play with it (search something benign or play trivia) then we'll have it on. It's good for that kind of thing, and those are useful enough we don't mind letting Google have the little data that might produce.
In any case having the capability to spy on you is fundamentally different from having the actual on-paper intent of doing so, such as a patent.
To the other point, I'm not entirely sure I'm comfortable with the patents, but it's at least an Echo simply gleaning more from a voluntary engagement with it. I think the general feeling is an Echo is "spying" when it engages itself to listen to your home.
Of course, I have no way to know if my echo is doing that, or if it were to start, and it's out of my hands to prevent it. The same with a smartphone.
If Amazon wanted to spy, probably a smart tactic would be to deliberately "flaw" the wake word recognition such that it could plausibly be a bug, but the result is activating very often randomly. They already kind of do that, it would just need to be more frequent.
One can check if a device is compromised or not behaving as expected fairly easily by monitoring network traffic.
Disclaimer: I could be completely wrong, and/or the situation could change over time due to improved hardware and software. There was a rumor earlier this year that the Google Assistant might add custom hotword support in the future:
https://9to5google.com/2018/01/29/google-app-7-20-assistant-...
Main source: Apple’s article about “Hey Siri”:
Still doesn't change the fact that placement of the Alexa is stationary, so your phone (regardless of how low the probabilities are) is still a considerably more "high risk" device in terms of what it could pick up than an Alexa.
To me this just sounds like another person in a whole line of them that screams about privacy unless they have to give up something that's extremely valuable to them. The Alexa isn't valuable to you. Your phone sure is.
At the base of the claim that Amazon is listening to what your echo can hear is the assumption that they're lying, and at the base of the claim that your iphone isn't listening to you is the assumption that Apple is not.
Personally, I wouldn't put it past either company to massively violate my trust if the incentive stars aligned just right, and I think it's silly to pretend otherwise.
The problems of putting a surveillance device in your home and carrying one in your pocket are roughly equivalent, and the rest is a debate about which huge collection of people with a wide array of beliefs and ethics is gooder
A more likely path is law enforcement's habit of putting them in the PR hell of "we could have saved Little Tiny Baby X if these companies weren't so hell bent on protecting perverts and terrorists". Once you have the device, it's only a matter of someone making a sympathetic enough public case that they need access to it.
On that front, I think Apple is taking more proactive steps to keep themselves out of the situation, I'm just not sure it's possible to avoid it completely.