After Facebook announced a new audio recognition feature for mobile users last month, many became concerned for the tool’s “Big Brother” potential. (A popular petition was even started to block the site from launching the feature.) That’s why Facebook responded this week in attempt to assuage user concerns.

Gregg Stefancik, Facebook’s head of security infrastructure, spoke to reporters in Australia about what the new feature can do — and, more importantly, what it can’t.

“The microphone doesn’t turn itself on, it will ask for permission,” he said. “It’s not always listening… so it’s very limited in what it is sampling. I wouldn’t want this in my pocket either if it was recording everything going on around me.”

He went on to say that the tool works by pulling a “audio fingerprint” from what users are watching or listening to and then comparing the sound to its existing database, meaning that it won’t be able to pull content from users’ potentially overheard conversations. He also noted that the app won’t store any audio data that it collects.

“The user is in complete control and the audio fingerprint that we’ve received is disposed of immediately,” Stefancik said. “The raw audio never leaves the phone and the data about the match is only stored if you choose to post it.”

Thankfully, since the feature is opt-in only, no user will have to use it if they don’t want to.