Can AI Solve for Diversity Representation in Marketing?

Representation Matters

With U.S. population growth being driven exclusively by minorities¹, diversity representation in media and advertising is more important than ever. Yet the major stock photography libraries – where most photos and illustrations are sourced for commercial or editorial use – are severely lacking in multicultural imagery. From portraits of diverse models (across race, ethnicity, age, body size, etc.) to lifestyle images of diverse family structures and inclusive themes and concepts, the limited availability of “diversity stock” has been a pain point felt by many in the industry for decades.

When I began my career in multicultural marketing in the late 90s, I remember struggling to find more than a handful of Black or Latino models (cast and shot in the U.S.) to use for campaign key art. And the ones that were available were often old or overused. Simply put, the people and the scenarios they depicted didn’t look at all like the world I lived in. From stale shots of faceless businessmen shaking hands to (overly) effervescent, grey-haired couples walking down the beach, not only did most stock photos lack diversity, they also lacked the natural authenticity marketers seek for their consumer campaigns.

Diversity in Stock Photography

To be fair, however, GettyImages – the largest player in the space – has made strides in this regard over the years. Through acquisitions and intentional work with diverse creators (from across the globe) they have dramatically expanded their libraries. But we’re not there yet. Frustration continues to grow – not only over the quantity of diverse stock, but also the quality and context of the images themselves. Often, scenarios depict tokenized minorities and/or reinforce stereotypes and other social biases.

In view of this critical gap in the market, smaller, minority-owned image libraries have emerged to challenge the stale status-quo of stock. Among them, sites like TONL, Nappy, and CreateHER Stock offer a broad selection of premium stock photos featuring “melanated” models in traditional and non-traditional scenarios.

However, with the rapid growth of Artificial Intelligence tools like ChatGPT and AI-powered digital image generation tools like MidJourney and Dall-E, it begs the question – Can AI images bridge the diversity gap in stock photography once and for all?

Inclusivity Through AI

AI-driven image generation technology can generate original digital images with infinite variations and customizations. This means that marketers can now access a virtually infinite library of diverse lifestyle and portrait photography and illustrations by using culturally-specific prompts in their search query. By leveraging this technology, creative teams are able to produce higher quality images faster than ever before, while still maintaining control over the content they are producing.

Ultimately, the use of AI allows individual creators and large brands alike to produce vast libraries of inclusive, royalty-free photos for content-driven marketing campaigns more quickly and efficiently than otherwise possible via popular stock photography sites or even original photo shoots.

While AI-powered image generation tools have the potential to make diversity representation in the advertising and marketing industry easier than ever, it still requires humans to key in the right prompts – to ensure the most authentic, inclusive, and representative output possible. So basically just like in most things in life, what you get back depends on what you put in.

If you’re a photographer, designer, copywriter, or model in 2023, you’re right to be nervous. The robots are quite literally taking over! But while AI technology is evolving quickly, there are still many limitations keeping it from becoming the new primary source for stock photos and illustrations. Photo-realistic images can still be (relatively easily) detected by the naked eye. And just recently, Getty Images announced it was suing a popular AI art tool for allegedly copying and processing photos from its copyrighted library².

That said, even with its flaws and limitations, the disruptive potential of AI is undeniable. It could very soon revolutionize the way we approach concepting, designing, and sourcing commercial imagery, let alone how we approach content development and brand communications all together.

Like all technology advancements, let’s just hope AI serves to evolve us into a more inclusive, tolerant, and equitable society. Too big an ask? Here’s to hoping.

###

Sources: ¹ Pew Research Center / ² CNN

Note: The images above are digital portraits developed using MidJourney through a complex array of cultural and photographic prompts.