Companies commit to mitigating AI bias in advertising technology

IBM and Ad Council research shows bias exists in advertising campaigns using data but says AI and technology can also address the challenge

A collection of organisations including technology vendors, tier-one brands and agency holding groups have committed to action aimed at mitigating bias in advertising technology.

The latest participants were highlighted during this week’s Cannes Lions International Festival of Creativity 2022 and include Delta Air Lines, IBM, WPP, Mindshare, 4A’s, IAB and the Ad Council.

As part of the collective effort, IBM and The Ad Council reiterated initial findings from a research initiative they conducted last year to explore whether bias does exist in ad technology. The research sought to identify unintended bias in algorithms and data from a creative message testing effort in The Ad Council’s Covid-19 awareness campaign, ‘It’s Up to You’. The campaign ran in late spring 2021 and focused on driving awareness of available vaccine options. As part of the research, 10 million impressions and over 108 different creative variations delivered via predictive dynamic creative optimisation were analysed.

The initial research findings released in January found confirmed bias was indeed in play, occurring around education, gender and income and driven by determined and proxy characteristics in advertising campaign data. However, it also showed mitigating bias in adtech was possible using artificial intelligence (AI) tools and resources through the marketing process. Off the back of this, IBM said it was clear more industry participation and data collection was needed to truly comprehend how much bias is impacting marketing campaigns.

As part of getting behind IBM’s efforts to take action, partners are signing up to IBM's ‘Advertising Fairness Pledge’. This asks signatories to commit to getting the facts on bias in adtech, explore the impact of bias in advertising campaigns and to be an advocate for advertising fairness.

Other brands who’ve signed on include Kellogg’s, MediaMath, dentsu, Publicis Groupe and The Female Quotient.  

“While the risk of bias in advertising is well known, by making this commitment, these organisations are among the first in the industry to take action,” said IBM Senior vice-president of The Weather Company and Alliances, Bob Lord. “Together, we are agreeing to educate ourselves and our companies and ask other industry leaders to join us in helping to mitigate bias in advertising.”

IBM has also released the ‘Advertising Toolkit for AI Fairness 360’, an open-source solution deploying 75 fairness metrics and 13 algorithms to help identify and mitigate biases in discrete data sets. A playbook and sample code are on offer so organisations utilising the toolkit can gain a better understanding of the presence and impact of bias on their ad campaigns, as well as makeup of their audiences.

“As a global brand, we know that every decision we make, whether it's about a supplier, an employee or an ad campaign, is a reflection of our values and the change we want to see in the world,” said Delta Air Lines managing director of brand marketing, Emmakate Young. “We've long been focused on inclusive representation in our campaign creative, this effort allows us to go a step further to bring more inclusive representation to our campaign delivery.”

WPP CEO, Mark Read, noted data can help brands personalise consumer engagement and identify relevant touchpoints – when used correctly.

“However, we know that bias can exist in algorithms or technology, and that's why we're helping our clients to evaluate how and when to use data in a meaningful way that will benefit the customer experience,” he continued.

WPP's GroupM has developed its own Data Ethics Compass for clients. Available as a global web app, this uses proprietary scoring logic to help advertisers evaluate the ethical risk level of data assets and decisions informed by them.

“IBM's new Advertising Toolkit for AI Fairness 360 will help us to better understand the potential impact of bias. Consumers rightly expect brands to use their information in a fair way and for the industry to tackle data bias collectively, which can ultimately result in increased engagement and commercial outcomes,” Read said.

Mindshare global CEO, Adam Gerhart, said it’s the agency’s duty to address bias in advertising head-on as technology and data prevalence accelerates.

“We believe the industry needs to take clear and intentional action, which is why we are committing to leverage the Advertising Toolkit for AI Fairness 360,” he said.

The IAB’s AI Standards Working Group also recently released a comprehensive report, Understanding Bias in AI for Marketing: A Comprehensive Guide to Avoiding Negative Consequences with Artificial Intelligence last November exploring the issue and what brands can do.

In the paper, the IAB noted several data and tech-oriented factors and challenges that could impact AI bias including data volume, data quality, data privacy and security risks, computing power, plus legal, regulatory, public and reputational risks. It also included checklists for advertisers when it comes to AI projects and utilisation, from discovery through to testing, activation, optimisation, remediation.

Read more:

Don’t miss out on the wealth of insight and content provided by CMO A/NZ and sign up to our weekly CMO Digest newsletters and information services here.  

You can also follow CMO on Twitter: @CMOAustralia, take part in the CMO conversation on LinkedIn: CMO ANZ, follow our regular updates via CMO Australia's Linkedin company page