Opinion: Beware of AI in Social Media Advertising - 2018

In a 2018 opinion piece for The New York Times, Farhad Manjoo warned advertisers about the growing reliance on artificial intelligence in social media advertising. Manjoo highlighted concerns about AI's ability to make biased decisions, lack of transparency, and potential for manipulation. This article delves into these issues, providing specific examples and data to illustrate the challenges and implications for marketers.

Analysis

Farhad Manjoo's 2018 New York Times opinion piece raised critical concerns about the use of artificial intelligence (AI) in social media advertising. Manjoo argued that AI algorithms, while powerful, are not immune to bias and can make decisions that harm users and advertisers alike.

One significant concern Manjoo raised was the potential for AI to perpetuate existing biases. For instance, a 2016 study by ProPublica found that COMPAS, an AI system used to predict recidivism in criminal defendants, showed racial bias, incorrectly labeling black defendants as higher risk than white defendants. While COMPAS is not directly related to advertising, it underscores the potential for AI systems to encode and amplify biases present in their training data.

In the context of social media advertising, AI algorithms are often used to target ads based on user data. If these algorithms are trained on biased data, they may disproportionately target certain demographics, leading to unfair or ineffective ad campaigns. For example, a 2017 study by the University of Southern California found that Facebook's ad delivery system allowed advertisers to exclude users based on zip codes associated with predominantly African American neighborhoods, effectively discriminating against potential customers.

Another issue Manjoo highlighted was the lack of transparency in AI decision-making. AI algorithms often operate as "black boxes," making it difficult for advertisers to understand how decisions are being made. This lack of transparency can lead to unpredictable outcomes and make it challenging to hold advertisers accountable for the content of their ads. For instance, in 2018, The Guardian reported that Facebook allowed advertisers to target users based on sensitive attributes like political affiliation, even though these attributes were not explicitly collected by the platform.

Manjoo also warned about the potential for AI to be manipulated by bad actors. In 2016, the Federal Trade Commission (FTC) fined Facebook $5 billion for allowing Cambridge Analytica to harvest the data of 87 million users without their consent. This data was then used to target political ads, highlighting the risks of relying on AI systems that can be exploited for malicious purposes.

For advertisers, these issues pose significant risks. Biased AI algorithms can lead to ineffective ad campaigns, while lack of transparency can result in reputational damage and regulatory scrutiny. Additionally, the potential for manipulation can undermine the integrity of advertising platforms, leading to decreased user trust and engagement.

To mitigate these risks, advertisers should adopt a cautious approach to AI in social media advertising. This includes regularly auditing AI systems for bias, demanding transparency from advertising platforms, and implementing robust data governance practices to protect user data. By taking these steps, advertisers can harness the benefits of AI while minimizing the associated risks.

A 2016 study by ProPublica found that COMPAS, an AI system used to predict recidivism, showed racial bias, incorrectly labeling black defendants as higher risk than white defendants.

The AdRes view

At AdRes, we understand the complexities and challenges of integrating AI into advertising strategies. Our suite of tools, including Prometheus for campaign planning, Odin for budget allocation, Athena for creative performance prediction, and Indra for real-time analytics, are designed to provide marketers with the insights and control they need to navigate the evolving landscape of AI in advertising. By leveraging these tools, advertisers can make more informed decisions, ensure transparency, and mitigate the risks associated with AI-driven advertising.

The takeaway

The integration of AI in social media advertising presents both opportunities and challenges. By being aware of the potential pitfalls, such as bias, lack of transparency, and manipulation, advertisers can take proactive steps to mitigate these risks. Regularly auditing AI systems, demanding transparency, and implementing robust data governance practices are essential for responsible AI use in advertising.