VCreatives Agency

How Generative AI is Changing the Face of Beauty and What the Beauty Industry Can Really Use It For

Generative AI has become pervasive in our reality. While the beauty industry seems to be embracing the new possibilities, we can’t underestimate the complexity of the interplay between AI technology, authenticity, and consumer trust.

Key Takeaways:

The beauty industry is increasingly relying on generative AI algorithms to create visuals, raising concerns about authenticity.
Attempts to marry AI with the evolving trends of diversity and inclusivity is proving to be quite tricky and leads to new complications.
Overreliance on AI-generated visuals in the beauty sector risks jeopardizing its credibility, potentially alienating consumers and eroding trust.
Though AI can enhance the customer experience, brands need to focus on authentic human bodies and foster relatability.

The AI Revolution

While we may not always realize it, over the years, we've gotten used to artificial intelligence in pretty much every nook and cranny of our lives. It’s quickly come to feel natural that voice-controlled assistants help find stuff online and recommendation engines suggest what to buy in e-commerce or watch on streaming.

We're cool with machines following a set of rules and churning out good solutions.

But can we be equally cool with generative AI algorithms that get creative with mere prompts? Will machines replace human creativity anytime soon? Or at least: Will leveraging AI for content creation unavoidably lead to less of the human touch in this content?

And just how indispensable that human ingredient really is? What benefits and risks can AI-powered innovation bring to beauty brands? And which of these will prevail? That's the puzzle the beauty industry is tackling right now.

AI in Beauty and Beauty in AI

The entry barrier to AI generative tools such as Midjourney or Bard is relatively low, allowing almost anyone to craft lifelike characters simply by entering prompts. Using natural language processing to transform words into visuals with an AI system can really sometimes feel like unraveling a bit of magic.

It happens in a flash, costs way less than employing human work and the outcomes are simply stunning. Ironically, that's where the twist comes in - it turns out, making AI images of people without making them look forever young and drop-dead gorgeous is quite the challenge.

Data-Driven Beauty Ideals

Delving into the roots of the issue: generative AI is a collective term for machine learning algorithms. They learn by soaking up data they are presented with. And the data they have been mostly fed so far – images showcasing peak human perfection in art and photography – has set a standard of idealized beauty for them to replicate.

In other words: we fed AI perfection, and now it's spitting out a world full of flawless humans.

Biases in Generative Algorithms

Because these beauty standards are also riddled with biases and stereotypes, using generative AI to produce images of people often results in outcomes that verge on caricature.

Request green eyes, and you might find them paired with "unexpected" red hair. Seek an image of a girl, and she's likely to conform to a size 34 or 36 standard. Specify a confident, intelligent, or dynamic-looking man, and you end up with someone sporting a body fat percentage rivaling bodybuilders.

It's a touch exaggerated, but you get the idea. Why should that be a problem for the beauty industry?

What about Diversity and Inclusivity?

Maybe we wouldn't even consider this an issue if it wasn't for the relatively recent push for diversity and inclusivity. After all, the fashion and beauty industries have for decades shown mainly one type of beauty, leaving out people of different backgrounds, body shapes, and sizes.

However, ever since 1989, when Naomi Campbell appeared of the Vogue cover, the issue of diversity has slowly but successfully made its way into the mainstream consciousness. In 2004, Dove launched its Real Beauty campaign, aiming to redefine traditional beauty standards.

In 2017 Rihanna's makeup brand Fenty Beauty made another small revolution by offering an unprecedented wide range of foundation tones.

And then in 2020 Gucci launched its "Unconventional Beauty" beauty campaign and engaged for it Ellie Goldstein, a model with Down's syndrome.

These and similar milestones have made representing diverse human bodies a must in both the fashion and beauty industry, as (especially younger) people have grown to expect this.

AI-generated Models and Influencers

Artificial vs Real Diversity

In an attempt to meet these expectations, fashion and beauty brands have already started to consciously include more diversity also in their AI-generated visuals. For instance, in 2023, Levi's turned to black AI models to promote inclusivity in their imagery.

At first glance, this might seem like a positive step. However, some argue that by opting for AI-generated instead of real models of color, companies like Levi's will now monetize diversity while avoiding any financial investment in those previously underrepresented groups.

This raises concerns about ownership and the loss of authentic representation.

Also AI influencers often depict individuals with diverse ethnic backgrounds. Take the "Balmain Army," for example. Introduced by the brand in 2018, it features representatives of three distinct racial backgrounds: Margot, representing Caucasians; Shudu, representing Africans; and Zhi, representing Asians.

But do they really represent anything apart from the perceptions of their creators? It seems artificial intelligence can only create artificial diversity that fails to give voice to the groups that it supposedly includes. And in the case of influencers, it additionally distorts the original purpose of social platforms - self-expression.

After all, representation goes beyond appearances; it's about empowering real individuals who diverge from conventional beauty norms to share their stories and challenge the prevailing narrative.

Compare the above examples with the Dove photo project, #ShowUs. Dove teamed up with Getty Images and Girlgaze, an online agency focused on diversity and inclusion, to build the world's largest photo library featuring 179 women from 39 countries.

All the photos were taken by female and non-binary photographers and are now available for media and advertising use. This is what real diversity looks like.

Not Just a Moral Dilemma

Even if you overlook moral concerns (we don't!), recklessly using AI-generated models and influencers to promote your products can have negative consequences from a business standpoint.

AI-generated entities advertising products designed for real people may simply seem too unrealistic to connect with, at best. At worst, your AI models and influencers could plunge into the uncanny valley, presenting an eerie and unsettling experience for viewers.

The Uncanny Valley Effect

The idea of the uncanny valley was coined by roboticist Masahiro Mori in 1970. Mori noticed that as robots became more human-like, they became more likable.

However, when they became too human-like, yet not quite human after closer examination, they seemed to fall into this "valley," leaving people with a feeling of unease or even creepiness. Since Mori first described this phenomenon, researchers have been spotting uncanny valleys everywhere.

There’s even a related theory, called perceptual mismatch, which says that we get uneasy when we notice discrepancies in features, like realistic eyes paired with unrealistic skin.

Does that sound like a challenge for beauty companies wanting to leverage generative AI tools for image creation?

Integrating Imperfections for more Authenticity

One way to avoid this effect is to make those computer-made bodies look more real by giving them some imperfections. Because, let's face it, being real means having a few flaws.

One savvy move from the playbook of a graphic designer is to use the deepfake face swap technique, in which you blend the flawlessly perfect face of an AI model with the real one of an actual human being.

Deepfakes and trust

However, also the deepfake technique comes with its fair share of challenges, mainly the risk of creating fake content passed off as genuine.

Big names in the camera business like Nikon, Sony, and Canon have already started incorporating tamper-resistant digital watermark technology into their cameras. This feature allows users to add a digital signature to their photos, acting as proof that their snapshots are the real deal, not some AI mischief.

This brings us to the crucial subject of trust.

Beauty and Skincare Brands' Credibility Dilemma

The beauty industry's association with trust has consistently been intricate. After all, it displays flawless beauty, portraying it as attainable for imperfect consumers (provided they use specific beauty products).

The utilization of unrealistically perfect AI images can jeopardize the credibility of a given beauty brand even more than traditional retouching. This impact may not be immediately apparent. The initial allure of flawless faces and bodies may captivate the audience.

However, the question arises: will this audience be able to relate to them? And, as a result: will the beauty brands manage to establish an authentic, subconscious connection with the real people they are attempting to reach?

The Toll of Flawless Faces

A couple of years back, the fashion industry caught heat for pushing unrealistic beauty standards with super-thin models (think Kate Moss and her BMI of 15!).

To tackle this, some countries put their foot down to look out for the health of models and young girls idolizing them. For instance, in 2015, the French National Assembly passed a law that bans the hiring of excessively thin models.

AI's Role in Daily Beauty Ideals

Should we maybe set similar limits on AI systems use? It's not just AI-generated models and deepfake techniques that pose problems. AI can also go the other way: rather than merging a flawless face of an AI model with the imperfections of a real human, you can apply AI filters to real faces and correct any flaws.

Apps like FaceTune and Snapchat have made photo editing easy and available to anyone, setting beauty standards sky-high.

In the not that distant past, when beauty ideals were defined by real models, movie stars and celebrities, the gap between them and regular people was clear. Now, with FaceTune and other apps, everyone can change their face shape or otherwise tweak their look, making unrealistic beauty norms a daily encounter.

All your friends and foes look perfect on social media posts. Constant exposure to flawless faces can take a toll on self-esteem and even lead to dysmorphia.

Leverage AI but Focus on Humans

The beauty industry revolves around human beings, their physical bodies and their imperfect beauty. To effectively navigate this industry, you must engage with real bodies because, ultimately, it’s real bodies you work for.

When showcasing products designed for people, you need to show how they interact with people. If the goal is to enhance the beauty of bodies, the initial step is to present them as they are, approached with tenderness and care to foster trust.

Some Beauty Brands Do it Right

That being said, it doesn't rule out the use of artificial intelligence. Here are some examples of beauty brands that successfully use AI to incorporate new ideas and transform their marketing strategies:

Glossier: Turning Customers into Influencers

Glossier is a beauty brand that heavily relies on user-generated content and word-of-mouth marketing. It uses AI to analyze customer feedback from reviews and social media posts to identify the most influential and loyal customers.

It then rewards them with free products, discounts, and invitations to exclusive events. This way, Glossier creates a community of brand advocates who spread the word about its products and values.

L’Oreal: Using AR to Personalize Customer Experience

L’Oreal, a global leader in the beauty industry, with a portfolio of over 30 brands, uses AI and augmented reality (AR) to offer customers a virtual try-on experience, where they can see how different products look on their skin, hair, and eyes.

L’Oreal also uses AI as virtual beauty assistants to create personalized recommendations based on customers’ preferences, skin and hair type, and beauty goals. This way, L’Oreal enhances customer satisfaction, loyalty, and trust.

Fenty Beauty: Employing AI to Promote True Inclusivity

As already said, Fenty Beauty is a beauty brand with the mission of catering to every skin tone and type. It uses AI to create and sell products that match the needs and preferences of its diverse customer base.

The brand's latest update is an AI-powered foundation shade finder. It not only transforms their customers' shopping experience but also showcases their dedication to inclusivity and innovation, and strengthens their brand ethos.

The Incredibly Important Question of Relatability

In navigating the integration of artificial intelligence into the beauty industry, beauty brands need to strike a delicate balance between leveraging AI's capabilities and maintaining authenticity.

The journey towards responsible AI implementation in beauty involves addressing concerns such as promoting impossible beauty ideals and navigating the complexities of diversity and inclusivity.

For successful brands it's crucial to use AI in a way that matches what especially younger customers have grown to expect: authenticity, naturalness, transparency, and inclusiveness.

Working with physicality demands showcasing authentic bodies, not their AI-generated counterparts.

It's about honesty that cultivates trust.

If you claim your skin care product will make someone's face glow, you need to show how it works on real people's skin in the course of real-life skincare routines. Presenting an unrealistic ideal in marketing campaigns hinders relatability.

If people can't in any way aspire to the portrayed ideal, they struggle to connect with the message, the product, and the brand.

‍