AI’s double-edged sword: Will it help or hinder marketing creativity?

Artificial intelligence has long been seen as the ‘next big thing’. But with ChatGPT’s rise to prominence, AI has suddenly moved along the hype cycle and become a potentially useful tool in numerous fields. That includes marketing, where brands are now experimenting with ways of creating value and saving money while preserving the human element of creativity.

If one thing is sure, it’s that the AI genie is out of the bottle, and this is a source of both anxiety and excitement for the general public. People are already wary of technology’s dominant role in their lives, Ipsos data shows, with 81% thinking big tech companies have too much power.

Perhaps even more startlingly, according to Ipsos UK head of creative excellence, Eleanor Thornton Firkin: “Sixty percent of people think that tech progress is destroying their life. ‘Destroying’ is a big word, so you can imagine that, as AI comes on top as well, it’s only going to add to today’s fear.”

However, people also see potential for AI to make their lives easier, expecting it to improve areas such as learning, entertainment and travel. Half even think it could improve areas such as their relationships and income.

These insights kicked off two roundtable events for marketing leaders, hosted in Cannes by global production house Tag in partnership with Marketing Week and its sister consultancy Oystercatchers. And as Thornton Firkin told attendees, the key lesson for companies using AI is that there must be room to keep an “underlying care and humanity at the core of what we do”, and to retain the “magic” in marketing that connects brands’ communications to people.

Marketing automation – and beyond

This balance is at the crux of how brands are deploying AI today. Ideas and content need to be generated at a faster rate than ever to meet business demand, but marketers and agency partners must learn how to do it with a positive effect on human-led creativity.

AI’s appeal in the near future is all about automation, as Toby Horry, global brand and content director at travel company TUI, explained: “We’ve got thousands of hotels, thousands of images and bits of copy, and having humans create that, keep on top of it and change it is time-consuming and boring, to be frank. So, for us, the first place we really start with using AI is more on volume tasks than creative tasks, because that’s where it’s going to make the biggest difference to us.”

When I’m going to need zillions of assets, that’s the bit I don’t want to spend loads of money on.

Sara Holt, Merlin Entertainments

Using AI to scale up asset creation can offer huge savings on digital ad production, as Sara Holt, group marketing director for the UK and Europe at Merlin Entertainments, revealed when describing a recent pitch process around the delivery of high-volume data-driven ads. She explained how one agency from a large holding company proposed a complex integration of data systems to meet the brief, while a smaller agency proposed an AI-based option at less than half the price.

When it comes to simply creating more assets, Holt said: “That’s exactly the answer you want to hear as a client. You don’t necessarily want to know that your agencies are doing anything differently, you just want to know it’s going to cost a lot less. When I’m going to need zillions of assets, that’s the bit I don’t want to spend loads of money on.”

Done right, this kind of production automation allows brands and agencies to direct more of their human resources away from repetitive tasks and towards creative work that adds more value. But AI can also help them find the spark that sets off the creative process. According to Hubspot, 33% of marketers who use AI use it to generate ideas or inspiration for marketing content.

There are also potential benefits for scaling organisations efficiently in multiple other contexts. For Darra Gordon, deputy president and COO of GLAAD, a non-profit focused on LGBTQ advocacy, AI’s potential lies in “the ability to grow the organisation with limited resources; to be able to do the work, use insights and research, and reach people” who would previously have been unreachable on its budget.

Experimentation: the promise and pitfalls

AI’s expansion from marketing automation into creative processes is only just under way, so the full extent of its benefits and risks is still being discovered. Richard Denney, joint chief creative officer at St Luke’s, said creating sample visuals is one example of where the agency is experimenting for presentations and mood boards, but within certain boundaries.

“We’re using it as a source of inspiration and to realise a potential visual you can’t find, but actually, it’s not the end result. And it shouldn’t be the end result. That’s where I wouldn’t allow it because you are stealing someone else’s intellectual property. If you say, ‘I want to make this image in the style of Rankin,’ and use it as final artwork, then you’ve stolen Rankin’s imagination.”

Richard Pinder, CEO of Rankin Creative – the agency set up by the eponymous photographer and co-founder of Dazed magazine – perhaps best sums up the general attitude to AI in business: “We are excited for our company, and slightly fearful for our world.”

He sees huge potential for AI to help get more done – for example performing colour correction on materials used in pitches – as well as in combining human creativity with the computing power of machines. However, he is cautious of the legal pitfalls, jokingly suggesting AI was created by lawyers to give them more work.

Intellectual property is a key area of concern. The UK protects the copyright of computer-generated works with no human author for 50 years, assigning it to the person or organisation undertaking the work, but few other countries follow this model so ownership of AI-generated creative may be difficult to protect. New works created by AI may also bear similarities to existing copyrighted works used to train the tools, which opens up risks of plagiarism.

However, this shouldn’t dissuade companies from experimenting. Indeed, AI does not always increase risk, it can also help manage it.

In heavily regulated industries such as financial services, for example, AI will be able to assist marketers with legal compliance around mandatory product communications, which do not need to be original, only precise. Today, creating these communications still requires many people, however AI can reduce human involvement in these low-value copywriting tasks while also cutting down errors and allowing people more time to work on creative communications.

Creativity starts and ends with humans

This faith in the continuing necessity of human creativity is one of the most profound aspects of the AI debate. Though we are already seeing AI-generated tracks pop up on Spotify, to engage and inspire people creativity still requires a human touch.

As Tag’s global COO Deepti Velury stated: “The core of humanity is having beauty and imperfection together.” AI-generated images are “too perfect” by contrast. “It just strikes you as not real, and I think that’s when the core element of humanity will prevail.”

There was consensus in Cannes that AI-generated creative assets tend to look the same, and St Luke’s Denney pointed out that brands therefore “still need the creative input to have something that’s separate from everyone else”. He gave the example of an awards entry he recently judged where the AI-generated creative required 1,000 prompts to arrive at the final, approved output.

The core of humanity is having beauty and imperfection together.

Deepti Velury, Tag

It is just one indication that, where marketing creativity is concerned, AI is unlikely to get the answer right first time. Human judgement is still required to determine what will connect with customers in a relevant and authentic way. For now, perhaps the ideal scenario for AI’s deployment in the creative process is that described by Alison Hanrahan, director of sales development, marketing and brand at Kone, a global manufacturer of lifts and escalators.

The company’s aim is to “create an environment for AI that puts our brand values and our brand guidelines into place, but then gives those to local marketers so they can be authentic locally, increase their own creativity, but still be in line with the guidelines. That’s how we’re thinking about AI, to instil more creativity rather than take it away.”

Future questions for AI

It’s already hard to remember what some areas of life were like without AI – from photo apps to smart assistants. What’s more, we haven’t even imagined all its possible future uses yet, but even in the medium term there is obvious potential for huge tangible business gains. In the travel sector, for example, “imagine scheduling 60 aircraft across a year and the role that AI can play in that”, TUI’s Horry speculated.

But along with AI’s evolution must come greater scrutiny. Horry is particularly concerned about the opacity around its carbon footprint. “How much server space is this actually using? And what’s the sustainability?” he asked. “What’s the impact if, all of a sudden, the entire world starts using AI?”

According to Tag’s Velury, carbon emissions from AI are estimated to be double those of doing the same work manually. It’s therefore important to find an alternate renewable energy source to fuel AI’s growth. “If you are a listed company and you are going to be talking to your shareholders and investors about carbon emissions, we can’t even really calculate it yet because we can’t prove how many assets are being created.”

The effect on ESG efforts is one of several issues that brands will have to address in order to manage risks in the long term. Another is the unconscious bias of humans’ inputs into AI tools, the materials they learn from and the algorithms themselves, all of which could lead to marketing that lacks cultural sensitivity, authenticity and representation.

Yet ultimately AI’s full implications – both benefits and risks – will only become clear when it is put into practice at scale, in a wide variety of applications. And getting to that point requires companies to experiment.

As Velury argues: “Try it, but with governance. Get the rules in place quickly on what you can use it for. And make an investment to fail, to play around, to see what does and doesn’t work.”