Unstructured Data, Training and Copyright Threats Thwart Brands’ Adoption of Gen AI

The boom in generative artificial intelligence has brands grappling to benchmark themselves against others and struggling with how to move forward with tangible use cases.  

Staying abreast with the smorgasbord of gen AI tools in the market is one challenge. Brands yet to explore gen AI are confronted with a knowledge gap and uncertainty about where to begin. Those brand marketers who have begun experimenting with this technology are concerned about regulation and governance.

“The promise of these tools being effective and efficient—and finding new business value to provide to customers—is in the scaling of it,” said Ashley Wood, global principal, brand innovation & insights, Ogilvy Consulting. “It’s just as much a change management issue as it is a technology issue [within brands].”

An August survey of 900 marketers, the 2023 State of Marketing AI Report from Marketing AI Institute, found that 78% said their organizations lacked any form of AI-focused education or training. In a June study, WARC found that 64% of marketing organizations are either in the early stages of adopting AI or haven’t started implementing it at all.

According to a Gartner survey conducted during the Summer with 405 marketing leaders, 62% of the respondents said that their current focus on technology and talent demands significant resources, leaving little room to explore emerging technologies like generative AI. To that, 53% of those surveyed expressed feeling overwhelmed by their current technology and project commitments, making it challenging to explore gen AI.  

Running into governance issues

Ogilvy’s brand partners, particularly those in the healthcare sector, have expressed concerns regarding consumer data privacy and IP protection standards related to gen AI. One hospital client’s marketing team was shut down for concerns over compliance after the team tested ChatGPT. The intricate privacy risks associated with open-source gen AI tools like ChatGPT have prompted firms like Samsung to ban staff using the chat tool following leaks of sensitive data, reported Bloomberg in May.

For some brands, the limitations surrounding indemnification and content ownership associated with existing gen AI tools have continued to pose challenges.

“Midjourney is probably the best gen AI image platform,” said Adam Buhler, svp of creative technology at Digitas. “But we can’t use it because its terms and conditions and [master service] agreements do not meet our needs. They don’t allow protection of our work.”

Reams of unstructured data

Brands are realizing the messy nature of their data stores hinders their capacity to fully use gen AI to optimize campaigns.

Brand partners of Digitas have been exploring how to use their data sets with gen AI. Their objectives included gathering campaign insights, making predictions for future campaigns and ultimately, improving conversion rates.

However, “when you get down to execution, you find that much of the data in the repository lacks context and is amorphous,” said Buhler. “We’re finding that to make sense of all the unstructured data, we have to handhold the generative AI so extensively that it’s not worth it.”

Benchmarking clients

Ogilvy Consulting is gathering evidence from clients to better understand their knowledge gaps.

Over the next few weeks, the agency plans to survey all its brand partners to assess where they stand on the AI maturity curve, determine their levels of ambition and understand their budget constraints.

“We found that what clients want the most from us is to share what other clients are doing,” said Carla Hendra, chief executive, Ogilvy Consulting. “If we can set some baseline and learn where various brands are today—and measure that on a regular basis—we will be able to extract insights to drive strategy.”