Thousands of Amazon workers are listening in on Echo audio, report says [u]
In a bid to enhance its in-house Alexa voice recognition technology, Amazon reportedly employs thousands of people around the world to listen, transcribe and annotate audio clips recorded by Echo devices.
Citing sources who have worked on the project, Bloomberg reports Amazon tasks outside contractors, as well as full-time Amazon employees, to comb through snippets of audio from Echo devices to help train Alexa, the company's voice-enabled assistant.
The clips include both explicit Alexa commands and background conversations. Echo devices are constantly scanning audio for a trigger phrase -- "Alexa," "Echo" or "computer" -- which when deciphered activates the assistant and initiates a connection to Amazon's servers. Audio recordings begin shortly thereafter.
"By default, Echo devices are designed to detect only your chosen wake word (Alexa, Amazon, Computer or Echo). The device detects the wake word by identifying acoustic patterns that match the wake word. No audio is stored or sent to the cloud unless the device detects the wake word (or Alexa is activated by pressing a button)," an Amazon spokesperson told AppleInsider.
With Echo in homes and offices, workers expectedly hear users discussing people by name or run across customers talking about sensitive topics like banking information, the report said. When snippets contain personal information, workers are instructed to mark the clip as "critical data" and move on.
Amazon does not inform customers that recordings of Alexa conversations are heard by employees, though the company's website does note user requests are applied to "train our speech recognition and natural language understanding systems."
According to people who have worked on the project, the job is typically uninteresting, with one employee saying he was tasked with parsing mentions of "Taylor Swift." Other recordings border on what can be considered intrusion. Two workers said they heard what they believe to be an instance of sexual assault, the report said.
Amazon reportedly constructed policies to deal with distressing audio clips, but two employees who sought help when they stumbled upon such snippets were told it was not Amazon's responsibility to take action. Employees have taken to sharing unsettling -- or amusing -- clips on an internal chat system.
The report notes Amazon's review system strips identifying information like a user's full name and address from the clips, but leaves the customer's first name and product serial number intact.
"We take the security and privacy of our customers' personal information seriously," Amazon said in a statement to the publication. "We only annotate an extremely small sample of Alexa voice recordings in order [to] improve the customer experience. For example, this information helps us train our speech recognition and natural language understanding systems, so Alexa can better understand your requests, and ensure the service works well for everyone."
With human intervention, Amazon is able to teach Alexa's software to recognize and respond to certain words and phrases. Like most voice-recognition technology, Alexa is powered by artificial intelligence, but the system requires training by human operators, in this case workers based out of Boston, Costa Rica, India and Romania. These people can listen to up to 1,000 clips per day, the report said.
Apple, too, employs a human review process to improve Siri. In a security white paper (PDF link), Apple notes Siri saves voice recordings "so that the recognition system can utilize them to better understand the user's voice." The recordings are stripped of identifiable information, assigned a random device identifier and saved for six months, over which time the system can tap into the data for learning purposes. Following the six-month period, the identifier is erased and the clip is saved "for use by Apple in improving and developing Siri for up to two years."
Machine learning, both on-device and in the cloud, is also used to tweak Apple's voice recognition technology, including "Hey Siri" wake word pronunciation.
Updated with statement from Amazon.
Citing sources who have worked on the project, Bloomberg reports Amazon tasks outside contractors, as well as full-time Amazon employees, to comb through snippets of audio from Echo devices to help train Alexa, the company's voice-enabled assistant.
The clips include both explicit Alexa commands and background conversations. Echo devices are constantly scanning audio for a trigger phrase -- "Alexa," "Echo" or "computer" -- which when deciphered activates the assistant and initiates a connection to Amazon's servers. Audio recordings begin shortly thereafter.
"By default, Echo devices are designed to detect only your chosen wake word (Alexa, Amazon, Computer or Echo). The device detects the wake word by identifying acoustic patterns that match the wake word. No audio is stored or sent to the cloud unless the device detects the wake word (or Alexa is activated by pressing a button)," an Amazon spokesperson told AppleInsider.
With Echo in homes and offices, workers expectedly hear users discussing people by name or run across customers talking about sensitive topics like banking information, the report said. When snippets contain personal information, workers are instructed to mark the clip as "critical data" and move on.
Amazon does not inform customers that recordings of Alexa conversations are heard by employees, though the company's website does note user requests are applied to "train our speech recognition and natural language understanding systems."
According to people who have worked on the project, the job is typically uninteresting, with one employee saying he was tasked with parsing mentions of "Taylor Swift." Other recordings border on what can be considered intrusion. Two workers said they heard what they believe to be an instance of sexual assault, the report said.
Amazon reportedly constructed policies to deal with distressing audio clips, but two employees who sought help when they stumbled upon such snippets were told it was not Amazon's responsibility to take action. Employees have taken to sharing unsettling -- or amusing -- clips on an internal chat system.
The report notes Amazon's review system strips identifying information like a user's full name and address from the clips, but leaves the customer's first name and product serial number intact.
"We take the security and privacy of our customers' personal information seriously," Amazon said in a statement to the publication. "We only annotate an extremely small sample of Alexa voice recordings in order [to] improve the customer experience. For example, this information helps us train our speech recognition and natural language understanding systems, so Alexa can better understand your requests, and ensure the service works well for everyone."
With human intervention, Amazon is able to teach Alexa's software to recognize and respond to certain words and phrases. Like most voice-recognition technology, Alexa is powered by artificial intelligence, but the system requires training by human operators, in this case workers based out of Boston, Costa Rica, India and Romania. These people can listen to up to 1,000 clips per day, the report said.
Apple, too, employs a human review process to improve Siri. In a security white paper (PDF link), Apple notes Siri saves voice recordings "so that the recognition system can utilize them to better understand the user's voice." The recordings are stripped of identifiable information, assigned a random device identifier and saved for six months, over which time the system can tap into the data for learning purposes. Following the six-month period, the identifier is erased and the clip is saved "for use by Apple in improving and developing Siri for up to two years."
Machine learning, both on-device and in the cloud, is also used to tweak Apple's voice recognition technology, including "Hey Siri" wake word pronunciation.
Updated with statement from Amazon.
Comments
The only real risk to me is embarrassment, but it's not likely someone I know personally is ever going to hear me saying or doing something I'd wish they hadn't. It's true that I don't want an Amazon contractor hearing my passwords or financial codes, but in the absence of a way for them to determine exactly who I am, even that information is essentially useless to them.
In reality though no one cares what you think about it. People that don’t care about privacy will buy one, and those that do won’t. Pretty simple.
I am also glad that I do not use Android for my mobile OS, as I do conduct banking transactions from my devices and I do use them to make payments. I protect my privacy and I care about my finances.
Can you imagine some rabid liberal working at Google or Amazon or Facebook or some other place listening into my private conversations? Haha. Not a chance in hell that I would ever trust them or any extremists working there, as they have been busted so many times in the past for their unethical behavior. They would probably embark on a private vendetta against me, because my private conversations would most definitely trigger some snowflake asshat that was working there.
I do not trust them, I do not trust their intentions and I do not believe what they claim.
I’m curious if there is any similar wording in banking apps.
When people sign a deal with the devil, then they shouldn't be surprised that bad things might happen.
Speaking of banking information, if they can see your personal data then your account could be compromised. Take the Experian breach which took my SSN, addresses, and all my banking info for CCs, student loans, automobiles, mortgages, and who knows what else. For each bank they have even more specific data on purchases as well as bank accounts that are connected, and when I'm likely away from home based on spending habits which include locations This isn't something Alexa would record after hearing the wake word, but it's something countless people in many countries around the world probably have access to.
I'm certainly more concerned with that than an Amazon employee listening to a random audio clip where I'm asking about movies, the weather, to adjust lights, or music. It's a little embarrassing that they know I listened to Carly Rae Jepsen's Call Me Maybe on repeat for months, but I'll just continue to pay the ransom every month to Bezos so no one else finds out.
Well, excuse the hell out of me. I thought sharing opinions and observations was the whole point of having a forum in the first place.
What I worry about is web people think this maybe okay, then companies think this is pay for then to infringe on everyone's privacy. Three more people think this is okay, the worse it makes it for everyone else.
If you call XYZ Widgets to order a bucket of nails, you give a human being your name, address, telephone number, and credit card info. Why would you be more concerned about a having that call recorded by a glorified answering machine then having it translated by someone who doesn't know anything more about you than your first name? How is it any more invasive than any other method of doing business?
It doesn't matter to me whether you choose to do business via a smart speaker or not, but I don't understand your reaction to my considering doing so (I don't even have such a device at the moment). What's described in the article seems LESS risky than many actions we willingly undertake every day.
Okay. So what? Once you give permission, that's what happens. If you refuse, you don't get your banking done. I don't understand your point.
I neither have, nor want an Echo, but this isn't the big deal some people want it to be.