Deep fakes and the threat to terrorism prosecutions

They say a picture is worth a thousand words. If that is true, how much is a video worth? A million? We are inundated with images, both still and moving, wherever we go. We watch videos on our cellphones all the time, even if most of them are crazy cat ones. And given that anyone can make and upload their own creations to YouTube it is no exaggeration to state that we have never had so much choice on what to view, good, bad or somewhere in between.

Terrorist groups have not stood idly by during this video revolution. Entities such as Islamic State (IS) mastered the art of using images to spread their propaganda and to instill what they hoped was fear into the hearts of all of us in the ‘infidel’ West. These productions looked professional to me and were quite stunning in their use of imagery and sound. Hell, we even have our own Canadian jihadi video stars: both Andre Poulin of Timmins, Ontario and John Maguire of Ottawa were featured in IS propaganda.

You would think that when idiots like Maguire and Poulin willingly put their faces on material promoting a terrorist group like IS those of us in security intelligence, law enforcement and prosecuting teams would rejoice. What better evidence do we need to bring these extremists to justice? There they are, in full video, professing support for a listed entity and often calling for attacks in the West. Slam dunk, eh George Tenet?

Except that the videos may be fake.

In what was an astounding piece of fiction this year, American Jordan Peele issued a video in which it sure seems former US President Barack Obama called current US President Trump a ‘dipshit’. But it was Peele’s voice – apparently he is a good mimic – plus some inexpensive, widely available AI software that made it look like those words came from Obama’s mouth. And, according to Macleans’ columnist Stephen Maher, more powerful tools making the production of fakes that much easier is on its way.

What does this mean for our ability to prosecute terrorists based in part on video evidence? Will defence lawyers scream ‘fake video!’, thus getting the data thrown out of court? And will malefactors start producing videos of me praising IS as a way to blackmail or besmirch me? Are we heading to a world where no one know what is real and what is not?

In the cases of Maguire and Poulin it does not really matter because these two will never see the inside of a Canadian courtroom. Not because they are still in Syria or Iraq, or adept at not getting caught, but because they are both probably dead. That makes it easier for us in deciding what to do with some of the 200 or so Canadians believed to have left to join terrorist groups. But what of the others who do come home one day and whom we want to put on trial? We are already having a hard time prosecuting Abu Huzaifa Al Kanadi, the Canadian who told a New York Times reporter that he had killed for IS, before retracting his story in what I assume was a ‘holy shit’ moment (as in ‘Holy shit, I told a reporter what??’). Now factor in doubts about other evidence and see how successful we are.

At a CVE conference in Edmonton late last month, my US colleague Daveed Gartenstein-Ross gave a presentation on ‘deep fakes’ and, in response to a question from me on the impact on prosecutions, assured me that there are indeed ways to detect falsified images. I really hope so, for all our sakes.

By Phil Gurski

Phil Gurski is the President and CEO of Borealis Threat and Risk Consulting Ltd. Phil is a 32-year veteran of CSE and CSIS and the author of six books on terrorism.

Leave a Reply