What are deepfakes?

Deepfakes are a type of media generated using artificial intelligence, whereby images are combined to create new footage that depicts events, statements or actions that may never have actually happened. The end-product is not dissimilar to the kinds of computer-generated imagery (CGI) that has been a mainstay of Hollywood productions for decades, but deepfake technology is free, fast and easy to use and often requires no specialist equipment or training.

What is image-based Abuse?

Image-based abuse is when someone shares, or threatens to share, an intimate image or video of a person without their consent – whether the image or video is authentic or created using artificial intelligence.

Some scholars have argued that image-based abuse is baked into the very foundations of deepfake technology – indeed deepfakes are named after the reddit user “deepfake”, who in 2017 took the existing practice of non-consensually editing celebrity women’s faces onto pornographic content using photoshop and used artificial intelligence to similarly edit videos. While the original forum closed, there have been numerous alternative websites which have emerged that provide access to deepfake technology and deepfake generated intimate imagery. In recent times, so-called “nudify” apps have also proliferated – these allow anyone with a mobile phone to generate fake explicit images and videos of real people.

Image-based abuse remains the dominant use-case of deepfakes and current reports estimate that over 95% of deepfakes on the internet feature explicit imagery. Almost all of these are non-consensual and almost all of them target women.

Are there laws against deepfake abuse?

Yes, in Ireland deepfake abuse is treated the same as other forms of intimate image abuse and if found guilty, those who create or share (or threaten to share) such imagery can be imprisoned for up to seven years. More information is available here: https://www.gov.ie/en/department-of-justice-home-affairs-and-migration/publications/intimate-image-abuse/

Many other countries around the world are in the process of implementing similar laws against deepfake abuse, such as the US and the UK.

What else can be done to reduce deepfake abuse?

As well as passing laws and demanding better regulation by internet platforms, we can all help to reduce deepfake abuse by educating ourselves on the topic. Whether as an internet user, a bystander, a parent, or a teacher, most of us will encounter deepfake media in some form in the coming years. Understanding the very real harms of deepfake abuse can empower all of us to speak up and speak out when we see deepfake bullying, harassment or defamation. To our knowledge, the Deepfakes, Real Harms intervention is the first of its kind globally, aiming to encourage empathy and understanding so fewer people go on to engage in these behaviours at all.