• by Sean Hargrave , Staff Writer, January 7, 2020

Facebook has banned deep fakes, and that has to be a good move. The questions remains, though -- can it accurately keep to its word and isn't it all a little too late to try to put the genie back in the proverbial bottle?

The Facebook decision comes at the start of an electoral year for the United States, and shortly after Boris won a resounding victory in the UK election. The latter saw some extremely dodgy editing of BBCnews presenters talking about Labour's position on Brexit, which was edited to make it look as if their damning words were what they were saying, rather than coming from sources they were quoting.

It wasn't so much a deep fake as a piece of deceptive video editing. Facebook decided to take it down, but only because the Conservative Party was using footage it did not own the copyright for. It was a technicality on the intellectual property front, not a case of being banned for deceptive editing.

To get an idea of what the technology can do, it's worth watching this video of Corbyn and Boris endorsing each other. Although you know it's a joke, you can see where this technology could go as it becomes more convincing.

America has brought us both the jokes, such as Obama's image being used to warn of the dangers of deep fakes, as well as outright dangerous use of deep fake technology. The faked video of Democratic House Speaker Nancy Pelosi doctored to make her look as if she is drunk and slurring her words is probably the best example we have yet of this being used not as a joke, but as an attack tool. 

The question remains, then -- once these videos get out, is there much Facebook can do about them? The detail is a little sketchy on how they will be taken down or detected.

If it is a case of taking them down, then they will obviously have already inflicted a lot of damage by being published in the first place. 

It is a step in the right direction, but as we hear that AI is now empowering deep fake technology to improve at breakneck speed, it will soon prove difficult to spot the fakes. Even if Facebook bans them, they are becoming more sophisticated, and it is easy to see how the example of Nancy Pelosi will not be the last deliberate attempt to ridicule a political opponent.

Spotting and deleting them will prove ever more difficult, but these videos will thrive where there is no Facebook and Instagram to take them down. The technology is out of the box, and it is hard to see how it can be easily spotted and then regulated.