Truth in [AI] Advertising

With Black Kilt recently embarking on a new marketing journey, the influence of artificial intelligence (AI) on advertising and product messaging is at the forefront of my thoughts. Although this might not strike some as a traditional “security” topic, on the contrary, the impact advertising has on society and our behavioral patterns certainly keeps me up at night. In today’s musings, I will delve into the link between advertising and security, as well as the impact AI is having on the marketing industry.

I constantly hear from fellow business owners and pundits about how AI is writing ad copy and saving hours of manpower every week. A quick Internet search returned more than 30 advertisements and business listings for AI tools to assist with the ad creation process in just the first page of results. But, as a security professional with over a decade of experience in data protection, I know there’s a darker side to this seemingly innocent trend. In fact, a joint study by researchers from the University of Kansas and Florida State University suggests only about half of AI generated ads are openly identified as such. So why should consumers care, and why should businesses?

Traditional advertising in the US is required to follow specific guidelines set forth by the Federal Trade Commission (FTC). For companies using AI to generate their ads, if no significant human oversight is involved, they run the risk of violating a number of advertising requirements. These include advertising that is truthful and evidence based. In particular, current AI tools run the risk of generating copy that makes claims based on flawed, outdated, or debunked research. Violations of the FTC rules can result in civil penalties of up to $50,120 per violation. For smaller businesses, this can quickly add up to real trouble. And the risk extends from ad content to algorithm-based targeting. A recent example of ad bias based on computer algorithms is evidenced in Meta’s agreement to alter ad serving technology to settle a lawsuit brought by the Justice Department alleging discrimination in housing advertisements based on race, gender, religion and other characteristics.

Aside from the more obvious challenges AI can create when producing ad copy, there are some significant benefits for advertisers, which in turn may be concerning for their target customers. AI is capable of crunching massive amounts of data in ways traditional marketing hasn’t fully explored—or lacks the capacity to. As a result, it can create models to predict market openings, and even when and how to engage customers. A recent article from Entrepreneur explores these possibilities from a marketer’s perspective, outlining the benefits AI can provide.  As a counterpoint, ThinkML outlines the privacy concerns for consumers in a May 2023 article.  In short, AI may not know how to respect legal or ethical boundaries around personal data.  The result might be extremely effective ads, but at the loss of consumer confidence if discovered to be using data that was not authorized or knowingly provided.

One last concern with artificial intelligence in advertising is the opportunity to create completely false content or even deep fakes.  A silly example is this recent AI generated fake beer commercial. Although the ridiculousness of this commercial puts it over the top, it demonstrates what may soon be possible.  A Wall Street Journal article from fall of 2022 highlights several examples of ads seemingly featuring celebrity endorsements that were in fact computer generated, and never agreed to by the celebrities.  Although, thanks to the FTC rules outlined earlier, it’s unlikely mainstream ads will start using fake endorsements anytime soon, it’s more likely these will become more frequent on private social media groups and other Internet back channels that are much harder to police for inaccuracies.

There are certainly benefits for business owners to leverage artificial intelligence in marketing strategies and digital campaigns. The key takeaway from the preceding points is to not lose sight of the human element required as oversight. Those that blindly allow AI to create and manage ad campaigns with no oversight might find themselves embroiled in a lawsuit. In addition, new rules and regulations may shape what AI can and cannot do for businesses. It’s important for business owners and marketers to stay on top of new developments.

As for consumers, the outlook is not so bright. Expect a relentless barrage of well targeted, if not insidious, ads to ensue. The best approach is to be skeptical of pretty much everything, seeking alternate sources to validate claims and avoid biases. One may also want to dust off ad blockers and other technology to help curb the onslaught. The age-old adage applies now more than ever: if it seems too good to be true, it probably is.

Related Posts