Why Advertisers Must Use AI skillfully and responsibly

In September 2023, the Advertising Association (AA) established the ‘Artificial Intelligence (AI) Taskforce’ to develop a cohesive industry approach towards the tech. Co-chaired by Google and creative agency VCCP, (and drawn from the AA’s wider membership), the group has released its first output - a report that outlines the context and potential of AI in advertising, stressing the importance of responsible adoption.

Titled ‘Advertising and AI: Showcasing Applications and Responsible Use’, the paper features six case studies, highlighting AI applications in campaigns for brands like McDonald's, O2, and Belvita, showcasing innovations from AI-generated candidates for FIFA President to campaigns responsive to ambulance sirens.

A 12-point checklist for responsible AI adoption is also included, focusing on transparency, compliance, and ethics. As well as this, insights are offered from over 15 industry contributors into AI's challenges and opportunities.

“Advertising as a whole has not only been an early adopter of many of the advances made by AI, but also been a clear leader in applying the technology,” Yves Schwarzbart, advertising industry relations manager, Google EMEA, said. “Generative AI’s ability to let everyone tap into its hugely promising creative potential has perhaps been its most dramatic impact on the advertising industry to date and this report highlights some great work in this space.”

The AA’s paper also notes the transformative impact AI can have across the entire advertising value chain - from strategy and ideation to production, media execution, and regulatory compliance. This principle underpins VCCP's generative AI agency, Faith, launched last year.

“Used responsibly, AI will be an unparallelled accelerator of human creativity and imagination,” Alex Dalman, managing partner and head of social and innovation, VCCP London, added. “We hope our work with the AA’s AI Taskforce and this report will help with a more widespread and responsible adoption of AI – and hopefully inspire others embarking on their AI journey.”

Here Dalman discusses how brands are using AI across strategy, media planning, and creative work. She also reveals the challenges the industry is facing in responsible, sustainable, and ethical use of the technology.

How are you seeing brands best implement AI into their marketing?

AI has emerged as a transformative force in all aspects of the traditional advertising process. For strategy - there are interesting and innovative ways to gather insights like using synthetic focus groups. In terms of media - AI can assist with planning and real time evaluation. Whilst for creative work, AI tools are transforming how ideas are conceived and realised - it’s fantastic for concepting and bringing an idea to life in seconds.

There are efficiencies - which have seen the cost of creating assets come down significantly following an upfront effort to train models, which also is taking asset generation timelines down from days to minutes. But the really exciting developments are in using AI to do something new, rather than doing the same job quicker or cheaper, and these aren’t about efficiencies, these are about showcasing the effectiveness of creativity.

Some agencies and brands are doing this and stepping into creating usable work with AI. The difficulty here is gaining sufficient creative control over the outputs to ensure your vision is realised, as well as, making the outputs brand safe and consistent. Which always involves expert human input. Controlling the chaos is both a challenge and an opportunity for brands to really master the art of AI at the moment.

What are the major challenges for responsible, sustainable and ethical adoption of AI by the industry?

The primary challenge is the uncertainty caused by the very newness of generative AI. ChatGPT, Midjourney, etc, burst onto the scene in just a couple of years, and so nobody has long experience to draw on. We can make historical parallels with, for example, Photoshop and desktop publishing software, or with social media, or smartphones, or the Web itself, but even those didn’t race to adoption at this speed. That newness also manifests itself in the legal landscape; do old laws cover this new technology, or do we need new ones? Regulatory guidelines are out of date almost as soon as the first draft is published.

Then there are the definitions: what do we mean by responsible, sustainable, and ethical? Using synthetic people in virtual photoshoots leads to less work for human models, but facilitates the representation of more diverse physical characteristics, from body size to ethnicity. Which is more responsible? Much has been made of the energy required to train foundational AI models, and for each subsequent generation; but if a sequence of generated images can replace a photoshoot, with its associated travel and energy costs,which is more sustainable? In the end, each company will have to evaluate the options and come to its own decisions about what is responsible, sustainable, and ethical.

Is there one recurring mistake you see being made around the implementation of AI by marketers?

That AI/Gen-AI is free and easy to use. Yes things like ChatGPT are easy enough for most people to use, and seemingly ‘free’ (companies have invested a huge amount of money into training the models and at some point that money will need to be recouped), yet more complex tools like Stable Diffusion or Runway still require highly skilled and technical workforce to master them and make them actually useful for brands by controlling the output. They also come with a higher subscription price tag, and not to mention the server processing power needed to run them, especially if run in conjunction with others in complex workflows.

Our recent film ‘Finding Faith’ set out to see just how ‘easy’ and ‘cheap’ it seemingly is to make a Gen-AI film following a year of hype in the space - spoiler alert it’s not easy at all, and it also required a significant amount of human hours, not only to figure out a workflow of 15 different tools, but also from a craft perspective to get it up to scratch as a piece of film.

Another common mistake is marketers thinking of Gen-AI tools as shiny new toys that they are dying to use and therefore starting with the solution: “Sora looks amazing let’s use it on a new campaign”. Due to limited understanding of the tool’s capabilities, teams then tailor their ideas to the tools - as opposed to starting with a well-defined problem for which AI is one potential solution.

A final problem is a lack of transparency. There are lots of big promises and partnerships in the AI software landscape but these often feel like smoke and mirrors or black box solutions. The great thing about this report is that there is finally a collection of real, tangible work that is starting to be put out into the world which we can all learn lessons from.

What work is Faith doing in ensuring AI is boosting commercial creativity and not hindering it?

First of all it’s about transparency, being open and honest where you’ve used AI and also sharing with others what we’re learning in our work with Gen-AI.

Whilst there is an obvious place for the use of AI tools in boosting efficiency - in Faith we work hard to remember that, despite inboxes overflowing with EOP deadlines, creativity is rarely about efficiencies - it is about effectiveness (part of our DNA at VCCP).

We are currently working in two work streams at Faith. The first is changing how we do what we do anyway, building branded models that can generate usable brand images or consistent copy for example. The second looks at how AI can change what we do - with AI being the tool that unlocks an idea that’s impossible to achieve without it. We are looking forward to sharing some work from that second stream in the near future.

How does this report help marketers take into account issues around AI regulation?

The brief in short was about responsible AI use - the area is significantly still blurry so this is the most important time to be cautious. Whilst the report showcases how brands have been implementing AI - there is a focus on issues around self-regulation, sustainability, rights of artists and ethics.

The ‘Advertising and AI: Showcasing Applications and Responsible Use’ report can be found here.