Mattel and Google: a double standard for AI toys?

Mattel yesterday pulled the plug on Aristotle, a planned smart-speaker-cum-baby-monitor developed by the company’s Nabi unit. The product had generated controversy since it was announced in January, with lawmakers, pediatricians, and child advocates raising concerns about how the device would collect data on and influence the behavior of children. The Washington Post summed up the concerns:

For one, the existence of a home hub for kids raised questions about data privacy for a vulnerable population. It also triggered broader concerns about how quickly companies are marketing products to parents without understanding how technology could affect early childhood development.

Congressmen Edward Markey and Joe Barton fired off a letter to Mattel CEO (and former Google exec) Margaret Georgiadis last week suggesting that Aristotle raises “serious privacy concerns as Mattel can build an in-depth profile of children and their family. It appears that never before has a device had the capability to so intimately look into the life of a child.” The letter sparked a new round of criticism in the press, with Jezebel calling Aristotle “creepy as hell” and Buzzfeed quoting a child privacy advocate arguing, “We shouldn’t be using kids as AI experiments … If we don’t know what the effect is, then we shouldn’t be putting that in children’s bedrooms.”

The letter and its fallout seem to be what prompted Mattel to announce yesterday that it wouldn’t go forward with the device.

At the very same moment Mattel was killing off Aristotle, Google was promoting new “kid friendly” accounts for its line of Google Home smart speakers:

We’re making Google Home more fun for the whole family, with 50+ new experiences for you to try out. Learn something new, or imagine with storytime. There are also plenty of fun activities; go on an adventure with Mickey Mouse, identify your alter ego with the Justice League D.C. Super heroes, or play Freeze Dance in your living room. These experiences will be supported by Family Link accounts on the Assistant, letting parents create accounts for their children under 13.

Even infants and toddlers can now be registered for Google accounts, allowing the company’s AI chatbot, Assistant, to collect data on them, talk with them, and tailor experiences for them. “We automatically collect and store certain information about the services your child uses and how your child uses them,” Google notes, deep in its privacy policies, “like when your child saves a picture in Google Photos, enters a query in Google Search, creates a document in Google Drive, talks to the Google Assistant, or watches a video in YouTube Kids.”

It’s hard for me to see much difference between Aristotle and Google Home with Family Link. Both raise concerns about children’s privacy, both allow companies to develop in-depth profiles of kids and their families, and both entail “using kids as AI experiments” without any clear understanding of how their development will be affected. Yet while the press hammered Mattel, it treated the Google news as benign, if not praiseworthy. “Google is making Home better for families and kids,” declared TechCrunch. “Google Assistant will tell your kids a bedtime story,” wrote Engadget. Buzzfeed chirped:

Home is now more kid-friendly, too. It can understand the way kids talk better, and includes more kid-friendly games, like “Which fruit are you?”. New commands include: “Hey Google, let’s learn”, or “let’s play a game,” or “tell me a story.” Google is also partnering with Disney to create kid-first experiences.

When a toy company tries to put a listening device into a kid’s bedroom, it’s creepy. When a tech giant does the same thing, it’s cool.