As a Virginia lawmaker, Ibraheem Samirah has researched Internet privacy issues and discussed how technology companies can regulate the collection of personal information. Still, he was stunned to learn the full details of the information Amazon.com Inc. had collected about him.
The e-commerce giant had more than 1,000 contacts from its cell phone. There was a record of what part of the Quran Samirah, raised a Muslim, listened to on December 17th last year. The company was aware of every search he’d done on its platform, including one for books on “Progressive Community Organizing” and other sensitive health-related queries that he thought were private.
“Are you selling products or are you spying on ordinary people?” asked Samirah, a Democratic member of the Virginia House of Representatives.
Samirah was one of the few Virginia lawmakers to oppose an industry-friendly Amazon-drafted state privacy bill that was passed earlier this year. At the request of Reuters News, Samirah asked Amazon to disclose the data it had collected about him as a consumer.
The company collects a huge amount of information about its U.S. customers and began making this data available to anyone on demand early last year after trying to break a 2018 California measure that required such disclosure. (US Amazon customers can get their details by filling out a form on Amazon.com.)
Seven Reuters news reporters also received their Amazon files. The data shows the company’s ability to create strikingly intimate portraits of individual consumers.
Amazon collects data on consumers through its Alexa voice assistant, e-commerce marketplace, Kindle e-readers, Audible audiobooks, video and music platforms, home security cameras, and fitness trackers. Alexa-enabled devices capture people’s homes and ring surveillance cameras capture every visitor.
Such information can reveal a person’s height, weight, and health; their ethnicity (via clues contained in language data) and political inclinations; their reading and buying habits; their whereabouts on a specific day; and sometimes who they met.
A reporter’s dossier revealed that Amazon collected more than 90,000 Alexa recordings from family members between December 2017 and June 2021 – an average of around 70 daily. The recordings included details such as the names of the reporter’s young children and their favorite songs.
Amazon took the children on how to convince their parents to let them “play” and received detailed instructions from Alexa on how to convince their parents to buy them video games. Be well prepared, Alexa advised the children to spread common parenting arguments such as “too violent”, “too expensive” and “you are not good enough at school”. The information comes from a third-party program “wikiHow” used by Alexa, which, according to Amazon’s website, offers instructions from more than 180,000 articles.
Amazon claims not to own wikiHow, but Alexa sometimes responds to requests with information from websites.
Some recordings included conversations between family members using Alexa devices to communicate in different parts of the house. Multiple recordings showed children apologizing to their parents after being disciplined. Others picked up the children at the age of 7, 9 and 12 and asked Alexa questions about terms such as “pansexual”.
In one recording a child asks: “Alexa, what is a vagina?” In another: “Alexa, what does shackling mean?”
Amazon’s Alexa home speaker is available for sale in an Amazon retail store in San Diego, California, United States [File: Mike Blake/Reuters]
Little did the reporter know that Amazon saved the records before the company disclosed the data it was tracking about the family.
Amazon says its Alexa products are designed to record as little as possible, starting with the trigger word “Alexa” and stopping when the user’s command ends. However, the recordings of the reporter’s family sometimes captured lengthy conversations.
In a statement, Amazon said it has scientists and engineers working to improve the technology and avoid false triggers that cause a recording. The company warns customers that records will be saved when they set up Alexa accounts.
Amazon said it collects personal information to improve and customize products and services for individuals. When asked about Samirah’s recordings of hearing the Quran on Amazon’s audiobook service, Amazon said such data enables customers to pick up where they left off from a previous session.
The only way for customers to delete much of this personal information is to close their account, Amazon said. The company said that after the account is closed, it will store some information such as purchase history in order to comply with legal obligations.
According to Amazon, customers can adjust their settings for voice assistants and other services to limit the amount of data collected. Alexa users can, for example, prevent Amazon from saving their recordings or have them automatically deleted on a regular basis. And they can separate their contacts or calendars from their smart speaker devices if they don’t want to use Alexa’s call or appointment scheduling functions.
A customer can opt out of examining their Alexa recordings, but they must navigate through a series of menus and two warnings that say, “If you turn this off, speech recognition and new features may not work well for you.” When asked about the warnings, Amazon said consumers who limit data collection may not be able to personalize some features like music playback.
Amazon founder Jeff Bezos demonstrates the Kindle during an Amazon event in California, USA [File: Gus Ruelas/Reuters]
Samirah, 30, got an Amazon Alexa-enabled smart speaker during last year’s Christmas season. He said he only used it for three days before returning it after discovering it was collecting footage. “It really sketched me out,” he said.
The device had already collected all of its phone contacts, part of a feature that allows users to make phone calls through the device. According to Amazon, Alexa users must give the company permission to access phone contacts. Customers need to disable access to phone contacts and not just delete the Alexa app to delete the records from their Amazon account.
Samirah said he was also unnerved that Amazon had detailed records of its audiobook and Kindle reading sessions. Finding information about his listening to the Koran in his Amazon file, he made Samirah reflect on the history of the US police and intelligence agencies monitoring Muslims for suspected terrorist links following the September 11, 2001 attacks.
“Why do you need to know?” he asked. Samirah’s term ends in January after losing an offer for re-election earlier this year.
Sometimes law enforcement agencies request data on technology companies’ customers. Amazon announces that it is complying with search warrants and other lawful court orders that collect data that the company holds in an account while opposing “overarching or otherwise inappropriate requests.”
Amazon data for the three years ending June 2020, the latest available, shows the company has at least partially complied with 75% of subpoenas, search warrants, and other court orders requesting data on U.S. customers. The company fully complied with 38% of these requests.
Last year, Amazon stopped disclosing how often it complies with such requests. When asked why, Amazon said it expanded the scope of the US report to make it global and “streamlined” law enforcement investigation information from each country. The company said it has an obligation to adhere to “valid and binding orders” but aims to release “the minimum required by law”.
Amazon’s 3,500-word data protection declaration, which refers to more than 20 other pages on the subject of data protection and user settings, gives the company ample scope for data collection. According to Amazon, the guideline describes the collection, use and disclosure of data “in a manner that is easy for consumers to understand”.
This information can get very personal. Amazon’s Kindle e-readers, for example, closely track a user’s reading habits, as shown by another reporter’s Amazon data file. The disclosure included records of more than 3,700 reading sessions since 2017, including timestamped logs – down to the millisecond – of books read. Amazon also tracks highlighted or looked up words, turned pages, and promotions viewed.
For example, it showed that a family member read “The Mitchell Sisters: A Complete Romance Series” on August 8, 2020, from 4:52 to 7:36 pm and turned 428 pages.
Florian Schaub, data protection researcher at the University of Michigan, said companies are not always transparent about what they do with users’ data. “We have to trust that Amazon is doing the right thing,” he said, “instead of trusting that the data cannot be misused.”
