MarTechTalk: The risk and rewards of deploying AI in marketing communications

On 3 October, MarTech Talk host Kevin Craine was joined by Deyana Petrova, (Former) VP, Digital & Marketing, Trafalgar Travel; Nicola Meinders, Global Marketing Director, JONCKERS;
Henry Ajder, European Advisory Council- Meta Reality Labs, Meta; and Antony Cousins, Executive Director of AI Strategy, Cision.

Views on news

In a message posted to his 9.5 million Instagram followers, actor Tom Hanks said his image was used without his permission. “BEWARE!! There’s a video out there promoting some dental plan with an AI version of me. I have nothing to do with it.” Generative AI offers plenty of new opportunities in marketing, as well as challenges in terms of ethics and legislation. Synthetic impersonation, however, has been around for some time. Now we are at a point where biometric data is out in the public domain and is up for grabs and the latest GenAI models now need less data to create better outputs. Disclosure of the fact that some marketing content is AI-generated may have, however, implications regarding the efficacy of an advert. 

The opportunities that genAI presents for marketers

There are use cases where, whether AI has been used to create some content is irrelevant to the user. Drawing the line between cost-saving and scamming consumers is the marketer’s call. Segmentation, personalisation and increased efficiency are the low-hanging fruits of AI deployments. AI has a major role to play in creating content and will also impact how content is going to be consumed. Moreover, it will give momentum to providing content in the users’ first language too via machine translation (MT) engines, including cultural adaptations. However, it’s key to find the right MT engines for a particular use case and pair of languages to get good results. While a human must be in the loop for current genAI models to be trustworthy, what we’re expecting in the future is the emergence of AI autopilots. 

AI legislation is being prepared and enacted around the world – EU AI Act, UK’s white paper on AI legislation, White House commitment to disclosure and transparency. As a result, certain high-risk use cases will involve burdensome auditing and safety procedures. However, you can talk to your solution providers, who can help you make the right technological choices. 

The panel’s advice

Try to be as transparent about using genAI as possible. 

AI is a tool that can get into the hands of rogue users, therefore oversight and control are essential when leveraging these tools. 

Understanding and applying the right type of disclosure for different media is of utmost importance. It should be different for a TikTok video or a billboard. 

Make sure genAI can create real value for your business and that it resonates with your audience. 

To err on the side of caution, strive to overshare and be extra transparent to avoid getting penalised. 

Don’t spam at scale. Edit and personalise heavily. 

Clean up your data before starting a transformation project. 

Consider whether developing in-house or buying software products will have a better ROI.