A Facebook executive stepped outside of official channels of communication last night by tweeting about a negative rumor that seems to keep resurfacing no matter how many times the company denies it.
“I run ads product at Facebook. We don't - and have never - used your microphone for ads. Just not true,” tweeted Rob Goldman, vice president of ads products at Facebook. That includes Facebook-owned Instagram, he said.
Goldman was responding to a tweet from PJ Vogt, one of the co-hosts of the tech podcast Reply All, which is producing a segment about the persistent belief that Facebook spies on users through the microphone. Vogt had asked people to call in to share their stories of why they think Facebook may be using the microphone to collect information for advertisers. “Ideally we want people who have stories. Like, ‘I was talking about how I needed a new rocket ship. I saw an ad for a rocket ship,’” he wrote.
Facebook has repeatedly denied that it uses the phone’s microphone to eavesdrop for keywords that would prompt a timely ad. But users keep noticing ads on Instagram or Facebook that seem to surface just after a conversation about that very thing: blackhead removal, cat food, parkas, the Foo Fighters.
I was talking to my friend about how I need a phone holder in the bathroom bcause our counter is small and I got this ad an hour later pic.twitter.com/0GB5NH2kkc— megan (@MorganCrockett) October 26, 2017
Facebook claims these creepy coincidences are due to frequency illusion, the psychological phenomenon where you see something once and then it suddenly appears to be everywhere, and to the fact that a brand may be using “omni-channel marketing” which could mean the user was primed for the concept by another ad on TV, radio, or across the web.
Alternative conspiracies include the idea that your phone’s manufacturer is eavesdropping on you and reselling data to third parties, or that Facebook is using image detection to look at the photos you’ve taken.
I run ads product at Facebook. We don't - and have never - used your microphone for ads. Just not true.— Rob Goldman (@robjective) October 26, 2017
It’s not insane to imagine that advertisers would listen in over the microphone. In 2015, Motherboard wrote about how clips from voice assistants including Siri and Cortana end up being listened to by strangers. Identifying information is stripped, but it’s still disturbing: most users do not realize their voice commands, including texts — “I love you. Send” — are being recorded, let alone being listened to by human workers employed by a crowdsourcing platform. It also came out in 2015 that Samsung Smart TVs were surreptitiously recording audio and transmitting it to third parties.
The Facebook-is-listening-to-you story keeps popping up every few months even though the company keeps strongly denying it. We still have no satisfying answer for how you could be talking about a thing and then see an ad pop up, when you’re positive you never searched for it or discussed it in text. Facebook’s answer boils down to: Our mechanisms, combined with the mechanisms of other platforms, are so sophisticated that it has become difficult for most people to grasp how this association could be made without direct spying.
Facebook’s business is predicated on knowing its users really, really well, and using that knowledge to show them the ads they are most likely to take action on. It does a number of unsettling things, like showing you an ad for the bookshelf you were just looking at on the CB2 website, or serving you ads when your friend likes a brand’s page — “Sarah likes Target!” Why would users believe that tapping the phone’s mic would be over the line? Two things are clear: People find hyper-specific advertising disturbing, and people don’t trust Facebook not to be a creep.