Stephanie Pringle & Alison Fowler: Why Ai Generated Synthetic Faces Are Costing Brands Their Most Valuable Asset – Consumer Trust

By Stephanie Pringle (right) & Alison Fowler (left), Co-Founders and Casting Directors, Chicken and Chips Casting

As casting directors, traditionally when we are asked to quote commercial projects we do so based on the complexity of finding the actors or specific people that meet a certain brief, with the budget, media usage and competitor category taken into consideration. However, almost overnight, we have been met with an alarming increase in requests from producers and clients about including Ai usage in our talent quotes or to consent to the use of the Artist’s image to train Ai models or to consent for the use Ai to change an Artist’s performance without any consideration into what this means for casting directors and actors alike.

The question is, are marketers really considering how this request and the use of Ai to manipulate an actor’s performance or to replicate an actor’s image will impact consumer behaviour and brand trust?

Brands that generate faces with Ai are using models that are trained using scraped datasets of real human faces (actors, models and even just unsuspecting members of the general public) who have never consented to their image, likeness and other biometric data like behavioural traits including voice patterns or the way someone walks and physical traits like facial geometry, being used in the first place. No one has paid for the original data. Ai models do not generate the images from nothing. This is not innovation or ‘clever use of a tool in the toolkit’ (which we’ve seen a number of production companies defending recently) and it’s time we called it out for what it is – extraction and theft.

When you strip the real human element out of your storytelling, you also strip the story from the truth, thus, the relationship between a brand and their customer starts to fracture.

We’re currently facing a seismic shift in synthetic faces generated by Ai where it appears that overnight, brands decided that finding the truth in their brand storytelling didn’t matter anymore. And truth is at the crux of any kind of story telling. Without it, there is no story. We have entered into a “Perceived Era”, where reality is judged by a configuration of signals rather than absolute truth and given all of our rules and regulations around advertising standards it is absolutely absurd that Ai has been given the opportunity to get as far as it has in Australia.

When casting directors read a commercial script and brief, it’s not a process of logistics, we aren’t just matching a face to a role and ticking a demographic box. Not only is it necessary for us to understand the story the brand wants to tell and the audience or specific customer they want to connect with through the people we deliberately put forward to be cast in each frame, the art of casting is searching for the internal monologue (those nuanced micro expressions, subtle performances, comedic timing), selecting people with a specific quality or presence, testing the chemistry between actors, pairing of family resemblance and dynamics and the lived experience, authenticity and identity that an actor brings to a role. These are qualities that only a real human carries. These are the truths that signal to the audience that the story is real and they feel seen as a consumer which directly translates into brand trust.

Outcomes from recent research from YouGov and SmythOS are damning, the trust deficit is quantified by a growing rejection of synthetic communication with 62% of consumers trusting social media content less if they know it’s been generated by Ai and 52% of consumers report reduced engagement with content they suspect is Ai.

NielsenIQ research indicates that AI-generated ads fail to create the lasting impressions that human-led campaigns do. Without human intent, there is no connection, no emotional nuance and the content ends up being “forgettable slop” with absolutely no hope of any brand recall (or at least not the positive kind).

From where we sit, we can see a fundamental flaw where we have Ai operating on the “linguistic likelihood” (that is, a statistical calculation of what an image should look like based on a data set) while lacking the interpretive engagement that is required to convey the emotion or meaning behind the artist’s performance and the specificity of a real human’s presence on screen. The removal of human judgment at the centre of casting is a dangerous place to dabble in because if you are generating synthetic faces that are devoid of specificity this does the absolute opposite of what diversity and inclusion stands for – it’s systemic exclusion and it removes authenticity completely.

We get it, there are many ways Ai can support businesses, cost and process efficiencies, but efficiencies are a means, not the point. Just because it exists doesn’t mean it needs to be used for every part of the process, and if we do, at what cost?

Photographer: Rachel Yabsley

Register for FREE HERE and receive the Campaign Brief Daily Bulletin and/or the global Best Ads Best of the Week Bulletin.              Subscribe to Campaign Brief Magazine.

#More Creative News   #No paywalls   #No annual membership fees

Subscribe to Portfolio & Reel for current listings of Australian and NZ ad agency and production company leadership and key personnel