Credit: S&B Vonlanthen Via Unsplash
AI slop, that growing mountain of content gaming the system, seeking the attention of advertisers when it doesn’t deserve it, can be easy to spot if you know what to look for.
At its heart, AI lacks a human heart. The machine is just predicting the words that should come next. People add emotion, that's where the magic happens.
Slop often follows identifiable patterns such as repetitive formats, bot-like writing styles, poor engagement metrics and sites lacking editorial oversight, according to to Conrad Tallariti, SVP & managing director APAC at DoubleVerify, a software platform for digital media measurement, data and analytics.
Nick Grinberg, head of strategy at media agency Next&Co, says spotting slop is easy.
“You see it…and know what it is almost straight away - at least for anyone that has spent any amount of time flicking through socials,” he told AdNews.
“AI slop often lags real culture. It references tropes/memes in an inauthentic manner and just ends up looking and feeling awkward.
“Just to be clear, not all AI content is slop. In the right hands it’s a supercharger that helps creators move faster and gives smaller brands a real shot at competing. However, if a brand leans too hard into churning out AI content, it risks flattening its own personality.
“I think we may get to a point where we will value - and want to prove - that content was created by a human. Publishers, creators, and agencies who can prove human origins of their content may well end up commanding more attention from their content.”
Dave Mooney, head of planning, BCM, said AI has democratised the creative process, giving anybody with a laptop and a login the ability to prompt to their hearts content.
“But just because you can, doesn’t mean you should. When you increase access and lower barriers to creativity, you end up with an influx of AI pollution, AI garbage… AI slop.
“You can pick out slop a mile away when you see it, it’s cheap, generic and stinks of a lack of original thinking.
“Slop is a volume game, it’s all about churning out quantity over quality which has created an influx of ‘filler’ that slowly erodes our trust in the content we choose to spend our time with.
“Slop usually looks bland, feels repetitive or disposable and generally lacks a human perspective. It can appear too polished or uniform, trying too hard to be perfect, void of nuance or quirks that grab and hold your attention. This often leads to lower engagement – so no likes, comments, or shares? It could be slop."
Phil Ohren, founder & head of strategy at Intender, said AI writing often looks too perfect.
“Think overly tidy paragraphs, balanced sentence lengths, and transitions packed with ‘moreovers’ and ‘furthermores’.
“You’ll also see the same words crop up again and again – like ‘enhance’ – because the machine leans on safe, predictable language.
“Humans, on the other hand, naturally break rhythm. We throw in short sentences. Long meandering ones. Even the odd emoji . That uneven texture is a sign of genuine authorship.
“One easy tell is formatting. AI has a strange fondness for em dashes, horizontal rules (< hr >), and numbered lists – structures that feel a little mechanical when overused. It also avoids the messy bits of real writing: half-finished thoughts, sudden pivots, or conversational quirks.
“If the article you’re reading feels like it could have come from a textbook, check again. Chances are the fingerprints belong to a model, not a human.”
Alfie Lagos, founder and director, Lexlab, said there are, for now at least, a few easy ways to spot the slop.
“The long ‘em dash’ is a dead giveaway. Same goes for overly filtered or uncanny-valley photography, the kind that looks like it was scraped from a dream sequence.
“But the biggest clue is in the writing. It reads like the colour grey. No real point of view, nothing new or polarising, and rarely anything that feels like it was written by someone with skin in the game. Just words taking up space.
Angus Wright, website and SEO manager at Yango, says slop often lacks the polish and authenticity of human-created work.
“It's a low-quality, low-effort product that can be identified by a number of tell-tale signs. Look for unnatural patterns, inconsistencies, and generic, unpolished content.”
Common giveaways:
Repetitive and incoherent writing: The content may be rambling or overly structured, relying on stock phrases and repetitive language.
Unusual grammar: Excessive use of quotation marks or Em-dashes can be a sign.
Distorted or bland imagery: Images generated by AI often appear distorted or lack a distinctive style.
Mass production: Slop-generating sites frequently publish hundreds of articles a day across a broad range of topics, designed purely to rank on search engines for programmatic ad revenue.
June Cheung, head of JAPAC at Scope3, the collaborative sustainability platform decarbonising media and advertising, said AI slop is content generated purely to game attention algorithms and attract advertising dollars, usually at the lowest possible cost.
“Its purpose isn’t to inform, entertain, or add value, but to drive pageviews and monetise at scale,” Cheung said.
“The result is content that is shallow, repetitive, and often clearly not written by a credible human source.
“Consumers are getting better at spotting it and instinctively scroll away. But brands and agencies may be buying it unknowingly because AI slop often hides in the long tail of programmatic, bundled with Made-for-Advertising (MFA) sites.
“These sites typically have convoluted, opaque supply chains designed to maximise revenue and Scope3 data shows they emit roughly double the carbon of the market average.
“One of the most effective ways to detect this inventory is through emissions. High-emitting supply paths are a red flag that the media chain is built for monetisation, not meaningful content. AI-generated sites can be filtered out at scale using agentic tools and training agents to exclude non-trusted sources, or generic, low-quality content.”
Honor Severs, head of creative at Vonnimedia, said AI models operate on tokenisation. They break down language into discrete units: words, parts of words, or characters and then assign them a numerical value.
“The AI's genius lies in its ability to predict the most probable sequence of these tokens based on the vast data it was trained on,” said Severs.
“This process is efficient and formulaic, but it's an exercise in probability, not in intentional, artistic structure.”
Compare this to the language of Homer's The Odyssey. The poem's power comes not just from its words, but from the dactylic hexameter - its strict meter and rhythm.
Severs said the poem uses formulaic phrases, known as epithets (such as "rosy-fingered Dawn"), not out of convenience, but to create a specific rhythm and cadence.
“This repetition served a purpose in the oral tradition, making the story memorable and powerful. It was a conscious, creative choice that gave the work a soul,” he said.
“AI slop, on the other hand, mimics this repetition without the soul. It uses cliches and predictable phrasing not to create art, but because that's the statistically probable output.”
Have something to say on this? Share your views in the comments section below. Or if you have a news story or tip-off, drop us a line at adnews@yaffa.com.au
Sign up to the AdNews newsletter, like us on Facebook or follow us on Twitter for breaking stories and campaigns throughout the day.