Artificial intelligence did not invent hollow, interchangeable content built to game algorithms. But it makes that content much faster, cheaper and easier to produce. Turning "slop" into a trial of AI is picking the wrong fight. Yet dismissing AI's role would be naive: it radically changes the scale of the problem.
Seems like AI slop
In late July 2026, LinkedIn launched a "Seems like AI slop" reporting button. It sits in the three-dot menu (...) at the top right of every post or comment. LinkedIn officially states that this is a low-value signal sent by members, and that its purpose is not to judge how a piece of content was made.
The platform specifies that AI-assisted content remains welcome if it carries a genuine voice, experience or expertise. Conversely, generic, repetitive, recycled content with no point of view is less likely to be widely distributed.
In other words: the problem isn't AI. The problem is the absence of value. Yet we call it "AI slop", not simply "slop".
Bad content already had its own industry
Long before ChatGPT, the internet already knew how to produce emptiness at scale: SEO-optimized content farms, clickbait headlines, articles rewritten on an assembly line, slideshows artificially split into pages, fake reviews, copy-pasted content, posts engineered to trigger a reaction rather than convey an idea.
In 2009, Wired already described Demand Media as a virtual factory capable of publishing 4,000 articles and videos a day, on topics selected by an algorithm based on search queries, competition and keyword advertising value. Two years later, Google launched the Panda update to reduce the visibility of pages deemed shallow or low-quality; the change affected nearly 12% of U.S. search queries.
There was no generative AI involved. There was already an economy of volume, a dependence on distribution algorithms, and a permanent temptation: publish whatever captures attention at the lowest possible cost.
Slop wasn't born with the machine. It's born every time production logic overtakes editorial intent.
What AI actually changes
Yet dismissing AI's role would be naive. It radically changes the scale of the problem. A mediocre editorial calendar used to require time, a team and a budget. It can now be filled in a few minutes. A generic image can be replicated endlessly. A boilerplate comment can be generated under every post. A fake piece of content can become credible enough to spread before it's fact-checked.
AI acts as a multiplier. It increases production capacity without guaranteeing relevance, originality, accuracy or taste. When the starting point is weak, it industrializes weakness. When the starting point is strong, it can instead accelerate research, widen creative options, sharpen a phrasing or adapt one concept across several platforms.
Technology amplifies the intention behind it. That's why anti-AI bashing is the wrong lens: it confuses the tool with the editorial choice.
Why hunting for "AI style" leads to a dead end
By trying so hard to spot AI through its tics — overly clean sentences, overly symmetrical lists, em dashes, demonstrative phrasing, an excessively smooth tone — we end up punishing an aesthetic rather than a lack of substance. A human-written text can be bad. An AI-assisted text can be precise, personal and useful.
This obsession even produces a new pitfall: artificially adding typos, abrupt breaks or fabricated confessions to "sound human". That's just another form of algorithmic staging.
Slop is first and foremost a content strategy problem
Content becomes slop when it is published to fill space: because the calendar needs feeding, because a trend must be followed, because a set pace has to be met, or to move a volume metric. It can be produced by a human, by AI, or by both.
Its two telltale signs are simple: it's interchangeable — any brand could sign it — and it's inconsequential — the reader understands, feels or does nothing different after seeing it.
For brands, the point is not to use less AI on principle. It's to bring more rigor before, during and after generation.
Let's not wage war on machines. Let's wage war on indifference.
LinkedIn's button has at least one virtue: it makes a collective fatigue visible. But if we turn that fatigue into a trial of artificial intelligence, we'll miss the real problem. Platforms have always been shaped by the incentives they create: publish more, react faster, repeat what works, optimize what gets measured. AI only accelerates that mechanism.
The answer isn't to return to a golden age that never existed. It's to build an editorial discipline fit for the age of abundance: less filler, more point of view; less imitation, more proof.
A simple framework for using AI without producing slop
- Start from a conviction, not a prompt. If the brand has nothing specific to stand for, AI will only create an illusion of a message.
- Inject what the machine doesn't have. Proprietary facts, lived experience, hard trade-offs, culture, taste, accountability and risk-taking.
- Give AI an editorial function. Explore angles, synthesize sources that still need verifying, challenge a structure, produce variations or adapt a concept. Not decide alone what deserves to be said.
- Edit more, not just produce more. Check facts, cut generalities, choose a pace, commit to a voice, and reject content that adds nothing.
- Measure value, not output. Success isn't measured by the number of posts. It shows up in attention earned, useful conversations, memorability, preference and action.
In summary
AI didn't invent mediocre content: it industrialized its production. "Slop" refers to interchangeable content, published to fill space rather than convey an idea, with no real effect on the reader. The real challenge for brands isn't giving up AI, but giving it a clear editorial function — exploring, synthesizing, adapting — while keeping the decision, the proof and the point of view on the human side.
FAQ
What is AI slop?
AI slop refers to low-value AI-generated content: generic, repetitive, recycled or lacking a point of view, published at scale to fill space rather than provide useful information or perspective.
Is AI responsible for the poor quality of content on the internet?
No. Hollow content existed before generative AI, driven by content farms and volume-based logic optimized for search engines. AI only accelerates and industrializes an already old practice.
Why did LinkedIn launch a "Seems like AI slop" button?
To let members flag low-value content. LinkedIn specifies that the goal is not to judge how a piece of content was created, but its relevance and originality.
How can a brand use AI without producing slop?
By starting from a clear conviction, injecting proprietary facts and real experience, giving AI a role in exploration rather than decision-making, and editing more rather than simply producing more.
Should you avoid text that "sounds like AI"?
No. Punishing a style rather than a lack of substance pushes toward counterproductive artifice (added typos, fabricated breaks). What matters is the value of the content, not its surface aesthetic.
And you — does your content bring real value, or does it just fill space? Let's talk about it.