Eighty percent of brands have concerns about agency use of GenAI

As a WFA member, you can download a copy of the survey here. Please note that this survey is WFA member content only. If you'd like to know more about WFA membership, please contact membership@wfanet.org.

Nearly two-thirds of brand owners (63%) are already using generative AI in their marketing strategies, according to new research from the World Federation of Advertisers (WFA) but many have serious concerns about the technology.

Eighty percent of multinational brand owners have expressed concerns about how creative and media agency partners are using generative AI (gen AI) on their behalf and legal (66%), ethical (51%) and reputation (49%) risks were also cited as major roadblocks to more widespread adoption.

Marketers also admitted to a lack of capabilities in this area (55%) and a general lack of AI maturity (49%).

Despite these concerns, however, the use of gen AI is rising, with usage well ahead of the 45% recorded by WFA research last year. In 2024 only 9% of brands say they have no current plans to use gen AI for their marketing.

The research is being released at the same time as WFA voluntary guidance, developed with marketing consultancy R3, on the steps brands could consider taking to mitigate the legal risks of gen AI in agency contracts. It is designed to help the 55% of companies who currently plan to review their media and creative agency contracts to cover the specific risks associated with AI usage as well as its potential benefits.

The new guidance follows an earlier publication from the WFA on the most commonly adopted responsible AI principles, along with standardised definitions. 

The findings are based on 54 responses from 48 of the world’s biggest companies, with a cumulative annual marketing spend of $102bn.

Most brands are still at the early stages of their gen AI journey. Just over a quarter (26%) claim to be ‘aware’ (they use gen AI in sporadic, occasional way) and 58% claim to be at a ‘developing’ stage (they have a dedicated gen AI strategy but it does not apply across all business units/teams).

Twelve percent consider their organisations’ use of gen AI as ‘mature’ (they have a unified gen AI strategy that applies across all business units/teams), and a further two percent as ‘leading’ (gen AI is embedded in how they drive innovation) and ‘transformational’ (gen AI is changing their business models).

“To harness the undisputed potential of gen AI, brands must first up get a grip on the legal and compliance implications. WFA’s AI Community and the guidance it’s producing will supercharge companies’ abilities to navigate the challenges so that they can prioritise the opportunities.” said Stephan Loerke, WFA CEO.

Currently, just 36% of companies have introduced terms prescribing how their partners can use generative AI on their behalf, and only 29% have reviewed their media and creative contracts with partners to introduce AI-related clauses.

Nevertheless, 48% are planning to introduce such terms and over half are planning to review contracts (55%). The primary motivation behind plans is to ensure compliance with data governance, introduce warranties and indemnities and ensure ownership of outputs. Just nine percent of brand owners are reviewing contracts with creative talent (such as actors and voiceover artists), to provide for generative AI use in relation to their voice or likeness, but almost half (46%) are planning to.

"Ensuring that agency contracts are updated to address gen AI is an action every brand should take. Not only is it effective risk management, it's an opportunity for marketers and their agencies to initiate important conversations about the role of AI in their partnerships." said Greg Paull, Co-founder & Principal, R3.

Other findings from the research include:

• Generative AI is most commonly being used in the areas of content creation (79%), content ideation (67%) and task automation (54%).
• Seventy percent are prioritising the use of gen AI for efficiencies (time and cost savings) over marketing effectiveness (increase revenue and returns on investments). While a few claim that gen AI use is driving ‘better media ROI’, ’50-80% increased efficiency’ and ’40-50% time saving for low-added value tasks’, many claim that it’s still too early to measure the true impact.
• Brand owners remain cautious about using AI-generated content in marketing assets (e.g. for text, image or music generation). Only 40% of respondents are currently doing so, although 37% plan on doing so soon. Of the 40%, a majority are using generative AI to alter or enhance real places, images or products (68%), to generate product descriptions or marketing copy (68%) or to generate places, images and products (59%). Just over one in three are using gen AI to copy the image/likeness of existing places or products (32%).
• Brands remain cautious, however, and less than one in three of the 40% who are using gen AI in marketing assets are doing so for chatbot purposes or customer support (27%), to alter or enhance the appearance of talent (23%), or to create synthetic sounds (18%), humans (14%) or voices (9%). And only 5% are using gen AI to copy the image, voice or likeness of real talent. And only eight percent are partnering with AI-generated influencers.

To govern the use of generative AI, 63% of brands have already adopted responsible AI principles, while 21% are still in the process of developing them. The most adopted principles are privacy (78%), transparency (76%), responsibility (70%) and respect for intellectual property rights (65%).

However, only 35% of respondents have taken efforts to translate principles into market practice, with 40% saying they plan to in the near future. And less than half (44%) have policies or guidelines on the use of generative AI for marketing.

For those who have developed guidelines and policies, these are focused on limiting the use of generative AI for certain purposes and use cases, providing do’s and don’ts to teams on how they should or should not use generative AI, outline approved internal use cases for gen AI and limiting the use of certain tool providers.

Over half of brands (54%) also have an AI governance board in place, and 68% of those who have one say it addresses specifically the use of gen AI in marketing.

The survey results will guide the work of the WFA’s AI Community over the coming months.

For information about the AI Community, contact Gabrielle Robitaille. For all media enquiries, contact Will Gilroy.