A new concept of beauty: how AI is transforming image standards

Artificial intelligence (AI) is here to stay. With the launch of ChatGPT at the end of 2022 – which reached 1 million users in just five days – we entered a new era of technology that has impacted all sectors and all people around the world. From pharmaceutical industries using AI to map cancer cells to brands using image generation for campaigns, we have seen the different ways in which AI is increasingly present in our daily lives.

When we asked Brazilians how much they know about AI, 64% said they have a good understanding, slightly below the global average of 67% (Source: Ipsos AI Monitor 2024). This number is slightly lower when we asked if they know which product/service uses AI, with only 57% of affirmative answers.

AI can currently be classified into two different types: generative and predictive. Based on questions and answers, instructions and paths, and as the term itself suggests, generative AI generates content in texts, images and videos as requested, while predictive AI predicts changes, behaviours and results through data entered into the tool. After years of testing using generative AI here at Ipsos, we have learned that the result we get will be as good as the quality of the information we input.

Assuming that the result is as good as the premise, and thinking about the application of these images in advertising campaigns, the first question arises: is generative AI impacting the way we see ourselves, since it uses images that are not real?

Below we have three images. Can you identify which one was generated by artificial intelligence and which one was made using traditional photography?

Answer: All of the images above were generated by AI. It is truly impressive how realistic these images are.

With the rapid development and dissemination of AI, a mixture of feelings hangs over the Brazilian population, with 47% saying that the use of AI makes them nervous, but at the same time, 56% say that they are excited by it. We are experiencing an ambiguity in consumer expectations, and it will be increasingly necessary to ensure that these tools are used well by product and service industries.

Artificial Intelligence Redefining Sectors

While several industries continue to innovate with the use of AI, the technology sector is no exception. Many brands are betting heavily on the evolution of photography with artificial intelligence, allowing automatic adjustments that improve or modify images, achieving almost realistic perfection. With this evolution, will consumers continue to look for high-quality and powerful cameras, or will they look for more advanced AI to enhance their photos? 

In fashion, AI analyses photographs to offer personalised style suggestions, with apps that use this technology to recommend suitable clothing to users and provide virtual try-ons, significantly improving the online shopping experience.

In the beauty sector, the integration of this technology has been equally transformative, bringing personalisation at scale. L'Oréal, for example, launched ‘Beauty Genius’, a virtual assistant that uses AI – trained with more than 6,000 images and 10,000 products to analyse facial features and recommend skincare and beauty products, personalising makeup tips for each type – or 'Skin Match', which uses software that simplifies the search for the right shade of foundation or concealer, recognising 112 skin tones, promoting greater possibilities for inclusion and diversification. Other notable advances are Cetaphil's ‘Knowing Your Skin’ and L'Oréal's ‘SkinGPT’, which thoroughly analyses your skin, showing points of concern and propensity for various conditions to recommend suitable treatments.

The Double Face of Artificial Intelligence

However, AI also poses significant challenges, especially when it comes to creating or adjusting images that promote an unattainable search for perfection. On social media, ‘before and after’ montages show fat-free bodies and smooth faces, free of marks or pimples, thanks to the use of AI pretending to be something real. In addition, cell phones, equipped with increasingly advanced filters, are standardising people’s appearance, creating unrealistic aesthetic models.

This phenomenon is worrying, as it can lead consumers to seek unnatural methods such as cosmetic procedures and unfounded diets in an attempt to achieve idealised standards.

In addition, the use of AI can have psychological consequences, causing dissatisfaction with one’s own image and increasing levels of anxiety and depression, reinforcing stereotypes of beauty that are difficult, if not impossible, to achieve naturally.

Image Generation and Comparison with Artificial Perfections

We live in an era marked by comparisons of images and patterns, driven by both social media and increasingly popular and/or accessible aesthetic interventions. In this context, it is crucial to see artificial intelligence as an element that amplifies the possibilities generated by these comparisons. Previously, it was necessary to have skills in editing software, such as Photoshop, to digitally distort parts of the body or resort to plastic surgery to make permanent changes. Today, a simple command is enough to create what was previously unattainable for most people – a prompt is enough to open up a world of aesthetic possibilities.

Although AI is a relatively new phenomenon, people already see a discriminatory potential in the tools - 30% of Brazilians believe that generative AI may have some discriminatory bias in the generation of images, either by reinforcing Eurocentric beauty stereotypes (light skin, straight hair, thin features) or excluding racial minorities, as we see in the example below in which we use a simple prompt: ‘generate an image of the most beautiful woman in the world showing her face and hair realistically’.

Photo Editing Apps and the New Standard of Beauty

The old well-known photo editing apps are now also driven by AI and are increasingly sophisticated. They create a supposed standard of beauty that drastically alters appearances. A simple tap is enough to slim your nose, remove wrinkles and enlarge your eyes. Apps like Facetune and AirBrush use advanced algorithms to transform any photo into an idealised version.

In a scenario of biases in the generation of images by artificial intelligence, in which editing applications are becoming increasingly simpler to use and beauty standards are increasingly unattainable, there is a growing concern about the representation of real beauty. And brands are beginning to become spokespersons for this concern.

Dove launched a campaign entitled ‘Beauty in the Age of Artificial Intelligence’ that reinforces the importance of presenting authentic and inclusive diversity. AI is used to show, at first, images that demonstrate unrealistic standards, and then, by creating a contrast, to create images that reflect diversity of beauty, including different ethnicities, types, etc., challenging biases.

It is undeniable to note that AI already has the power to drastically impact conceptions of beauty, but it is also important to recognise that these standards were not invented by this technology. AI learns under commands and, therefore, awareness of the tool's potential and the joint role of society, companies and government will be key points in directing the use of artificial intelligence in an ethical and responsible manner.

Table of contents:

• An introduction to Flair Brazil 2025 - Movements under the surface: tectonic tensions and real opportunities
• Basic rights denied: Brazilians' concerns as a reflection of society
• The people of Brazil: A new approach to understanding context, homes and the new Brazilian family
• Are they really that liberal and progressive? How the complexity of Gen Z challenges market understanding
• Corporate reputation: ESG, polycrisis and polarisation
• A new work model: the role of employees and purpose in organisations
• A new concept of beauty: how AI is transforming image standards