Fifty percent of online content is AI-generated, according to a recent study from Graphite. With this type of content dominating everything from how-to-guides and news articles to blog posts, brands are finding their ads appearing alongside AI content for the first time.
In some cases, ads are appearing alongside AI ‘slop’ – low-quality, mass produced, GenAI images and videos – raising new questions about brand safety and the potential consequences of appearing alongside inappropriate user-generated content.
Appearing Next to AI Content
Misinformation or fake content can quickly damage brand credibility if an ad appears beside it. This has always been the case for content misaligned with ads. But now the average layperson has access to GenAI tools like ChatGPT, Claude and Runway, the influx of AI content online – much of which is low quality – makes it more challenging for brands to vet the content they’re advertising alongside.
Some websites act as ‘content farms’ designed to rack up ad impressions and take advantage of programmatic budgets, posing some very real challenges for advertisers.
Csaba Szabo, Managing Director EMEA at media optimisation platform Integral Ad Science, says the challenge for advertisers is distinguishing value-adding AI content from the kind of content that erodes trust and lowers ad effectiveness.
He says: “Signals like source credibility, content accuracy, article depth, and visual quality are key in recognising potentially low-quality, AI-generated content, along with plagiarised content, hallucinated facts, and filler phrases that add no value.
“To combat this risk, advertisers are shifting focus from surface-level metrics to true performance outcomes to avoid low-quality, AI-generated inventory. Greater supply chain transparency, along with adopting the latest AI-driven technologies to harness media quality metrics to optimise towards quality content, will be essential to safeguard brand reputation and drive greater outcomes for marketers’ budgets.”
Audiences Care Less Than We Think
A recent study from brand safety ad tech platform Zefr and Omnicom Media Group’s research arm OM Media Trials assessed the risks and opportunities of ads being placed next to AI content.
It found there’s no discernible difference in how ads perform next to AI-generated content, and in some cases ads perform well next to AI content. However, the report stressed the difference between AI content and AI slop, with not all AI content considered low quality and likely to damage brand perception. This highlights that it isn’t AI content in and of itself that damages brand reputation, but the quality of the content – something that has always been relevant.
Whether or not AI content is labelled also plays a significant role in consumer perception. The Zefr research found that 41 percent of consumers feel better about a brand when AI content is clearly labelled, with transparent use of AI tools improving brand perception.
When a consumer is unable to tell if AI was used to create content, and it isn’t labelled, brand performance declines across favourability, trust, and intent to purchase.
Rich Raddon, Co-CEO at Zefr told FutureWeek that it’s difficult for walled garden platforms, like YouTube and TikTok, to label AI content unless major GenAI firms “play ball”.
He said: “The platforms in general want to make sure AI content is labelled, but they’re a little restricted in terms of the scale of how much content is being uploaded to their platforms, the platforms rely heavily on uploader clarification. There’s an industry-wide initiative to try to put invisible watermarking on AI content, but this would rely on the foundational [AI] models.”
Despite consumer desires for AI content to be labeled, AI companies and social media platforms need to come together to watermark content. This, Raddon believes, isn’t necessarily up to platforms alone.
He continues: “It’s going to be challenging to determine and decipher [AI content]. But, I’m not sure the platform should play that role, candidly. In recent times, the platforms have really moved into the role of being the ‘public square’.”
