Facebook has announced that it will tackle the rise of problematic deepfakes on the site
At a time when video and photo manipulation tools are becoming more advanced than ever before, the social media giant has announced a policy change effectively banning deepfakes on the social network, fearing that they could be used to sway public opinion ahead of the forthcoming US election.
Monika Bickert, Facebook’s vice president for global policy management, said that deepfakes “present a significant challenge for our industry and society as their use increases.”
The blogpost, Enforcing Against Manipulated Media, was posted to Facebook’s Newsroom on Monday. It outlined how the social network will qualify, then remove “misleading manipulated videos”, based on specific criteria.
If the videos or photos are adjudged to have been either edited “beyond adjustments for clarity or quality” in a way that is sure to mislead somebody, or if the content is the “product of artificial intelligence or machine learning” in which modifications are made for it to appear authentic, Facebook will remove the content.
They were clear to point out, however, that this directive did not extend to parodical or satirical content, and creators would still be free to use the technology for entertainment so long as it adhered to the usual community standards.
Deepfakes could fall into the wrong hands
As the technology used to create deepfakes becomes more sophisticated and, perhaps more worryingly, more accessible, Facebook’s preemptive strike on what could be a deciding factor in the upcoming elections is welcomed by many.
Last year, a video of Barack Obama went viral. In the video, he calls President Donald Trump an obscenity, before highlighting the importance of verifying news on social media.
“This is a dangerous time,” he says. “Moving forward we need to be more vigilant with what we trust on the internet.”
However, this wasn’t the former president, but a deepfake public service announcement, voiced by American actor Jordan Peele.
As the screen splits, Obama’s mouth is synchronised with Peele’s impression, perfectly demonstrating how realistic, yet absolutely fake, a viral video can be, and its potential to disinform millions.
Facebook’s efforts to stymie the dissemination of manipulated media isn’t just for the removal of content. They also announced that employees, alongside partners from numerous cross-sector organisations and educational institutions, were working towards uncovering the identities of the creators, and learning how to better detect deepfakes in a bid to stop them at source.
“Just last month, we identified and removed a network using AI-generated photos to conceal their fake accounts. Our teams continue to proactively hunt for fake accounts and other coordinated inauthentic behavior.”