Kevin Long: My job could be done in a nanosecond with a computer program that we had in place before, and just because of the way that they’ve changed their rules after the Cambridge Analytica Scandal … all the networks readjusted their access point for the data. And it … made the job a lot harder.
Bob: Yeah, sure. Well, the criminals have automated tools, but you don’t.
Kevin Long: Right.
Last week, I spoke with you about the imposter problem on social media. Sometimes, it’s criminals pretending to be famous people via “cele-bait” — National Geographic photographer Kevin Nicklen shared harrowing stories with us about his decade fighting imposters. And sometimes they hijack regular folks’ accounts just to attack trusted friends. Either way, it’s a nightmare for multiple layers of victims. Meta says it removes hundreds of millions of inauthentic accounts from Facebook and Instagram, but victims tell me they’ve had nightmares trying to fix the …ongoing…problem. And as we discussed last week, there’s an entire company named Social Imposter that has spent the past 10 years working to protect brands from the stink of imposter incidents. The mere existence of this company means the problem is still severe.
This week, there’s more to learn from Social Imposter and its founder/owner, Kevin Long. I hope you’ll listen to this episode of The Perfect Scam, but I really want to make sure you hear about this terrible irony: How and why the Cambridge Analytica scandal actually made the imposter problem worse, at least according to Kevin. The tl;dr is this. Long’s company used to scrape Facebook and automate searches for look-alike accounts — think BobSullivan1234, BobSul1ivan, that kind of thing. Because Facebook was hit with legal action around the data sharing after the CA scandal, it tightly shut down third-party access to its data. So now, Long and his employees must manually search for imposters. Meanwhile, criminals highly automate the creation of lookalike accounts.
In the age of AI, only the criminals get to use automation. Here’s the relevant section of our conversation:
Kevin Long: So during the 2016 elections and the Cambridge Analytica Scandal hit, at that point I had a program that was using the APIs that the networks allowed me to do automated searches. So the searches that we had with the software that we built would contact Facebook, would contact Twitter, would contact YouTube, contact all of the networks and, and say, ‘Okay, I need a list of everybody named Kevin Long.’ And it would return all of the accounts that came back under that username.
When Cambridge Analytica hit, they had misused their access that they were granted through the APIs, and so Facebook, probably rightly so, did a review of everybody who had access to their APIs and shut it down during the review. So what do you do now? Now you’re not able to automatically search, now you’ve got to employ people to go in and do physical searches, manual searches, typing in names of your clients.
And so it went from being a job that was relatively easy to being one that was very time-consuming. And … it’s increased our workload tenfold because we have to do the searches manually now versus using a computer to go in and do it because once they… eventually they did reopen their APIs, but they changed the things you could ask for. So you could no longer ask for, “Bring me back everybody under the name Kevin Long,” which was all I needed. I didn’t need the stuff that you can ask for now. … It was irrelevant to what I needed to be able to perform my task.
So you know I spent several years trying to convince Facebook, “Hey, you’ve got to let me do this, because I’m helping you. I’m helping your community; I’m helping the victims in your community. This is not being used for nefarious purpose, this is being used for a very specific targeted purpose to help eliminate fraud.” And I just, it was spinning my wheels. I got nowhere with them. They weren’t going to make an exception for it… None of the other networks really seemed to care either or wanted to adjust either to allow me to do that. So now everything’s done manually.
Bob: So I’m sorry, but I’m picturing that a human being is sitting at a desk right now typing, Bob Sullivan, and occasionally substituting in a “1” for an L” or whatever … trying to imagine …. and they’re just brute force doing this one at a time?
Kevin Long: Yep, that’s how it’s done now. And that’s why it takes longer.
Bob: Oh my God. And this is the age of artificial intelligence.
Kevin Long: Yeah… my job could be done in a nanosecond with a computer program that we had in place before, and just because of the way that they’ve changed their rules after the Cambridge Analytica Scandal, it has … all the networks readjusted their access point for the data. And it … made the job a lot harder. It didn’t make it impossible; it just made it a lot harder and a lot less effective. But I mean effective from our perspective in that from a time perspective.
Bob: Yeah, sure. Well, the criminals have automated tools, but you don’t.
Kevin Long: Right.
To read the rest of the transcript from this episode, visit the home page for it by clicking here. You can also listen to the episode by clicking play below, or by clicking this link, or by searching for it on your favorite podcast platform.
Be the first to comment