AI and ad bias—how brands can develop ethical frameworks

By Sharb Farjami

Artificial Intelligence’s ability to process vast amounts of data and generate insights has the potential to transform advertising campaigns, maximize ROI and enhance brand growth while helping users be more productive.

But it is equally important to recognize and mitigate AI’s risks: Michigan State University was slammed for using ChatGPT to craft a student support email after a fatal campus shooting. “Tessa,” the National Eating Disorders Association’s helpline chatbot, was taken offline after sharing harmful weight loss advice. And there have been countless reports of brands using AI to post sham reviews for products and services.

As AI’s applications rapidly advance with much-needed regulation close at hand, the advertising community must agree on the steps to ensure ethical and safe implementation. Here are a few considerations for brand leaders: 

Avoid harmful misrepresentation

A recent magazine headline, “When is an image a harmful lie?”, highlights a challenge for AI users. While we enjoy memes such as the Pope dressed in Moncler, such an image raises concerns about potential harm. AI empowers creative exploration, but it is essential to exercise discretion. Brands once pledged against filtered or altered

images due to regulatory scrutiny. Today, distinguishing human and AI-created content is challenging. If AI is used in ad creative, consumer awareness matters. Clear guidelines, ethical boundaries, human oversight and transparent disclosure are vital for trust and authenticity in AI-generated content. 

Safeguard ad placements

AI-driven ad placement can harm a brand's reputation and lead to inappropriate or harmful ad contexts and brand safety issues. AI-automated placement, even with safety guardrails in place, might miss things that a human with a better understanding of context and culture could pick up on—hence the need for human oversight and audits to maintain fairness and brand safety. 

Prioritize authenticity

Over-optimization of advertising can result in ads that lose their authenticity and trustworthiness. Brands will need to strike a balance between AI-driven personalization and maintaining a genuine connection with the audience. It is important to leverage user feedback and to conduct regular sentiment analyses and refine AI algorithms to ensure you are coming to market in an authentic way.  

Ensure accurate ad targeting 

Inappropriately trained AI algorithms can mistakenly categorize users—creating false positives/negatives—resulting in poor ad targeting, cost inefficiencies and even potential harm to consumers. To improve AI accuracy, we need to prioritize transparency in data collection and use, allowing users to understand and control the information that they provide. Given that AI tools can be inherently biased, users must understand how the tools have been trained by their owners to mitigate against bias. We also need to ensure that all stakeholders—agencies, brand leaders and media partners—understand how to select input data to avoid a discriminatory output. 

Establish guidelines 

AI algorithms are complex. While clear disclosures about data and methodologies support ethical advertising decisions, AI complexity makes it difficult to make them accountably. We must develop robust ethical frameworks that address accountability, transparency and fairness. Implementing explainable AI algorithms and closely collaborating with legal, AI experts, media and marketing teams can help navigate ethical and legal dilemmas. 

Uphold ad quality standards 

When inexpertly used, AI-enabled technologies might contribute to low quality creative and ad clutter. We must emphasize quality control and moderation measures to maintain the overall integrity of the advertising ecosystem. Working with partners that implement content guidelines, provide user feedback mechanisms and proactively monitor ad output can reduce spam and improve ad quality. 

Set and protect ethical boundaries 

AI-assisted targeting techniques can push ethical boundaries by exploiting users' vulnerabilities. Brands will need to adopt responsible targeting practices, respecting privacy and avoiding manipulation. Implementing guidelines that prioritize informed consent, data anonymization and regular ethical reviews can ensure ethical micro-targeting and protect consumers. 

Keep sustainability top of mind 

AI activity, such as processing massive datasets through numerous iterations to optimize an AI model's parameters, requires a massive amount of energy. Given that so many brands are rightly committed to meeting aggressive sustainability goals, it’s important to consider how the use of AI in advertising might impact carbon calculator goals as well as broader sustainability initiatives. 

Check—and double-check—with legal 

Building generative AI-enabled content engines for digital advertising will revolutionize brand content and experiences. It will enable creative teams to produce high-quality commercial content faster, more efficiently and at scale while staying fully aligned with a client’s brand. In tandem it will be increasingly important to implement robust content creation guidelines for your brand upfront and review processes to ensure compliance with copyright laws and intellectual property rights.  

AI offers tremendous potential, but it also must make advertising better for people. This means wisely integrating AI technologies that help us meet our goal and drive improved productivity and disclosing its use. By embracing transparency, accountability and ethical guidelines, our industry can harness the power of AI technology to drive improved business outcomes while cultivating consumer trust. 

Sharb Farjami is CEO of Wavemaker North America.