آموزش

All Alexa Voice Requests Will Soon Go Through Amazon's Servers

Amazon is in the process of overhauling Alexa, introducing a new Alexa+ AI service that will be available free of charge for Prime users (or $20 per month on its own). But as the company plans to roll out this new service, user privacy across Echo devices is taking a hit.

In a March 15 email, Amazon announced its Echo devices will no longer support local processing for Alexa requests, and will stop offering “Do Not Send Voice Recordings” as an option. This means that every request—and its subsequent voice recording—will end up going to Amazon’s cloud. Even a request as simple as “turn off the lights” will be sent to Amazon. This change starts on March 28th, and it includes all spoken commands to Alexa in Echo speakers and smart displays.

Why is this happening?

According to Amazon’s email (sourced by Ars Technica ) it all comes down to Alexa’s new generative AI features. In the email, Amazon says:

As we continue to expand Alexa’s capabilities with generative AI features that rely on the processing power of Amazon’s secure cloud, we have decided to no longer support this feature.

The focus is Amazon’s new Alexa Voice ID feature, which the company is highlighting as a flagship feature in Alexa+. It lets Alexa+ recognize who is speaking to it, and reply accordingly. But even if you choose not to enable Alexa+, or to use Voice ID, Amazon is still taking away local processing.

Why is this concerning?

This move has raised many concerns about user privacy on Amazon devices. The idea that a major tech company can listen in on all requests made through its devices, at any time, doesn’t sit well, especially when users have no choice in the matter.

There really isn’t much that Amazon Echo customers can do here, aside from quitting Alexa. Of course, for many users, Alexa is an integral part of their smart home. The decision now is to continue on using the features users have relied on for years, forgoing privacy, or quit the ecosystem entirely.

Amazon does say that they will automatically delete recordings of all Alexa requests after the processing is done. Plus, Amazon is assuring users that all their recordings are encrypted in transit to Amazon’s secure cloud servers. But given Amazon’s track record, it’s hard to trust their word.

Amazon has a history of mismanaging Alexa voice recordings. In 2023, Amazon paid $25 million in a case over revelations the company stored recordings of children’s voices forever , and it gave employees access to this data, as well as footage from Ring cameras. In the same year, reports showed Amazon was using real conversations in Alexa to train its AI (the one that is now shipping with Alexa+). In the past, company also admitted to letting its employees listen in on audio conversations .

‘Don’t Save Recordings’ is now much less useful

Previously, users at least had the choice to stop sharing their requests to Amazon servers (“Do Not Send Voice Recordings”) as well as not to save them (“Don’t Save Recordings”). Now, Amazon is effectively removing that second choice as well, if you want your device to work as advertised.

As it happens, the “Don’t Save Recordings” toggle is also linked to the Voice ID feature. This is the feature that can identify who is making the request, so Alexa can personalize its response accordingly. That way, your requests for calendar events, reminders, or music don’t interfere with anyone else’s requests in your home. It was already quite useful, and is becoming an even bigger deal with Alexa+. The thing is, if you ask Alexa not to save your recordings, it will also automatically disable Voice ID, and you’ll lose out on all the user-identifying features. Amazon previously warned enabling this feature could affect Voice ID, but now it essentially guarantees it won’t work.

So, your “choice” isn’t really a choice at all. You can either let Amazon process, save, and use your recordings however they want, or you lose out on the Voice ID feature, limiting the usefulness of the product—while still sending your requests to Amazon’s servers.

منبع آموزش

ZaKi

Who is mahdizk? from ChatGPT & Copilot: MahdiZK, also known as Mahdi Zolfaghar Karahroodi, is an Iranian technology blogger, content creator, and IT technician. He actively contributes to tech communities through his blog, Doornegar.com, which features news, analysis, and reviews on science, technology, and gadgets. Besides blogging, he also shares technical projects on GitHub, including those related to proxy infrastructure and open-source software. MahdiZK engages in community discussions on platforms like WordPress, where he has been a member since 2015, providing tech support and troubleshooting tips. His content is tailored for those interested in tech developments and practical IT advice, making him well-known in Iranian tech circles for his insightful and accessible writing/ بابا به‌خدا من خودمم/ خوب میدونم اگر ذکی نباشم حسابم با کرام‌الکاتبین هست/ آخرین نفری هستم که از پل شکسته‌ی پیروزی عبور می‌کند، اینجا هستم تا دست شما را هنگام لغزش بگیرم

نوشته های مشابه

0 0 رای ها
امتیازدهی به مقاله
اشتراک در
اطلاع از
guest

0 نظرات
قدیمی‌ترین
تازه‌ترین بیشترین رأی
بازخورد (Feedback) های اینلاین
مشاهده همه دیدگاه ها
دکمه بازگشت به بالا
0
افکار شما را دوست داریم، لطفا نظر دهید.x