FKA(I) twigs

On 30 April, FKA twigs announced to the US Senate subcommittee that she is developing her own deepfake.

Ok, so what is a deepfake, you ask? Well, for those of you living under a rock, deepfakes are videos or sound recordings that replace someone’s face or voice, in a way that appears real.

Deepfakes present a unique challenge for copyright law from both an input and output perspective. For deepfake technology to replicate sound and vision accurately, the models are trained on copyrighted sound recordings and images, often without the owner’s permission. Even where deepfakes are used to create something entirely new, there are risks of emotional, financial and reputational damage for the depicted artist.

We recently saw this on Drake’s Taylor Made Freestyle, where he employs an AI-generated Tupac Shakur to diss Kendrick Lamar in their ongoing feud. Tupac’s estate was “deeply dismayed” at Drake’s effort to weaponise the deceased’s voice, and the track has since been removed after a cease-and-desist was sent to the Canadian rapper last week. This event seemed to be the tipping point for the committee, finally inviting discussions to endorse law and order around AI.

So, what has twigs got anything to do with it?

The British singer, songwriter and performer was part of a key witness panel to discuss the NO FAKES Act, which proposes to protect people from having their images, voices and likenesses used to create digital replicas that say or do things they never agreed to or would never say.

The Act aims to do this in two ways:

(1) Individuals and companies will be held liable if they produce unauthorised replicas

(2) Platforms will be held liable if they host or distribute unauthorised replicas knowing the person depicted did not authorise it

In the cellophane singer’s testimony, she emphasised how her music truly reflects her identity and stressed the importance of enacting regulations to safeguard the authenticity of artists and to “protect against misappropriation of our inalienable rights”. On behalf of all artists, she urged the committee to act quickly, learn from previous failed technology regulation responses and not make the same mistake with AI.

twigs added that she wasn’t against AI - just the misuse of it - and announced the development of her own deepfake, trained in her personality and tone of voice to speak many languages. She believes that by doing this, she’ll reach more fans and be able to explain the creative process of her upcoming album in the nuance of their language. Her deepfake will also assist with promotional duties so the singer can spend more time making art and embodying a “northern star” for her fans. Her innovative campaign for artist control over deepfakes showcases how AI can elevate artistry in a way that isn't harmful.

As long as deepfakes are regulated, permission can be granted more meaningfully and damage can be controlled. In theory, watermarking AI creations is one way to protect artists from the harmful effects of deepfakes, but it’s not entirely realistic. More secure solutions involve regularly updating legislation that punishes infringing users and/or facilitating an ‘opt-out’ mechanism for owners to block material usage.

So, will more artists follow in her footsteps? It’s uncertain whether artists will want to develop deepfake technology for themselves. However, FKA twigs is certainly leading an industry call for regulatory change. Jess Glynne, Zayn Malik and Sam Smith are just some of the artists who have recently teamed up to sign an international open letter supporting more protection from deepfakes.

And so the human war with technology continues…

Watch this space!