The way Facebook handles privacy has been called into question for years. Now, a developer has claimed that a third-party company is somehow scanning private Facebook messages.
His claim got a lot of attention this morning on Hacker News, a popular message board for software developers.
Except both Facebook and the company itself, Recorded Future, say they have no idea what’s going on.
Coder website Bosnadev wrote a blog post this morning looking into what happened when it posted a private link in Facebook chat. The group had built an app that had never been published. It then posted the link to the app in private Facebook chat box and tracked who opened the link.
For the first three minutes Facebook’s own IP address looked into the link. This shouldn’t come as a surprise — Facebook and most any other chat function routinely check links. But three minutes after that, an IP address from the company Recorded Future hit the link.
Recorded Future works to identify real-time online risks by collecting and analysing sources on the web. Recorded Future is a private company, as are most of its clients, but it does have ties with the government thanks to an investment from the CIA’s investment arm, which set the conspiracy theorists off.
But to us, the company explicitly denied looking at Facebook Chat messages.
“We have no idea
what the guy is referring to. We only get Public/Open Facebook data, and certainly have no interest nor any technical ability to get to any chat data at all,” Recorded Future told Business Insider.
Facebook also said it’s not working with Recorded Future, and wasn’t sure what the blogger has found.
“While investigating the claims of this post, we’ve confirmed that Facebook doesn’t use Recorded Future — an open source aggregator of public data — to scan any private content. That means we haven’t partnered with or directed Recorded Future to scan anyone’s message links. It’s hard to tell precisely what’s going on based on the amount of information in the post. It’s possible that another interaction, including one that could be occurring on the client machine, is consuming the URL and generating this behaviour.”
So what’s going on here? We’re not sure, but you can follow the discussion on Hacker News and see where it ends up.