Talk to us
Menu
← Insights

Insight · Problems with AI

AI Slop: How to Spot It in Your Own Marketing Before Buyers Do

Skein30 October 20267 min read

AI slop is low-quality content made with generative AI: generic, often wrong and not something anyone asked for. In business marketing it shows up as blog posts that could belong to any company, emails full of stock phrases and claims nobody checked. Spot it by asking whether a piece says anything only your business could say.

Key takeaways

  • Macquarie Dictionary named "AI slop" its 2025 Word of the Year, defining it as low-quality generative AI content, often containing errors.
  • Slop in business marketing is usually generic, unsourced and interchangeable with competitors' content.
  • The test is simple: could a competitor publish this unchanged? If yes, it is slop.
  • According to a 2024 GEO study, adding citations, statistics and quotations raised visibility in AI answers by up to 40%, so specific content also travels further.
  • The cure is specific input, not better prompting: facts, customer words, evidence and a human edit.

What is AI slop?

AI slop is content produced with generative AI that has little value: generic, padded, often inaccurate and not wanted by the reader. Macquarie Dictionary, Australia's national dictionary, made it the 2025 Word of the Year and defined it as low-quality content created by generative AI, often containing errors and not requested by the user.

The choice was unusual. Both Macquarie's committee and the public vote picked the same term, which shows how widely people now recognise it. Other dictionaries chose the shorter "slop" the same year.

For businesses, the danger is not the obvious junk on social media. It is the plausible blog post, newsletter or LinkedIn update that looks professional and says nothing. Readers increasingly have a word for it, and they skim past it.

Why does AI slop happen in business marketing?

Because AI is asked to write without anything specific to work from. Given a topic and no facts, customer insight or point of view, a model produces the average of everything it has read on that topic. Add pressure to publish more, little review and no evidence standard, and slop follows.

Three conditions produce most of it:

Volume targets. "Four posts a week" with no brief produces filler.

No input. The AI has no product facts, customer language or examples, so it writes what anyone could.

No review. Nobody checks claims or asks whether the piece is worth publishing.

How do you spot AI slop in your own content?

Read each piece and ask five questions. Could a competitor publish it unchanged? Does it contain a specific fact, example or customer insight from your business? Is every claim sourced? Does it use stock phrases and hype words? Would your best salesperson send it to a prospect? A "no" to the last question is decisive.

A quick slop checklist:

  • Interchangeable. Swap your company name for a competitor's; if nothing else needs changing, it is slop.
  • No evidence. No numbers, no named sources, no real examples.
  • Stock phrasing. "In today's fast-paced world", "unlock your potential", "seamless", "game-changer".
  • Padding. Long introductions that restate the title before saying anything.
  • Unchecked claims. Statistics or features that nobody verified.
  • Machine punctuation. Marks your own writing never uses, such as long dashes scattered throughout.

What does AI slop cost a brand?

It costs attention, trust and visibility. Readers learn to skip your content, prospects notice generic or wrong claims, and search engines and AI assistants favour specific, sourced material. Every slop piece also crowds out the content that would have earned a buyer's interest.

Visibility is the measurable cost. According to the 2024 study that introduced generative engine optimisation, adding citations, statistics and quotations from credible sources raised visibility in AI-generated answers by up to 40%. Slop does the opposite: it gives AI assistants nothing distinctive to cite.

AI assistants also lean on outside evidence. According to a 2025 University of Toronto study, AI search engines show a strong bias towards earned media over content brands publish about themselves. A brand blog full of generic posts adds little to that picture.

Clicks are scarcer, too. According to Pew Research Center, Google users clicked a traditional result in 8% of visits when an AI Overview appeared, against 15% without. When fewer readers click through, the content they do read needs to be worth it.

How do you stop publishing AI slop?

Feed AI specific material before it writes, brief each piece with a real angle, require a source for every claim, and have a person review against a short checklist before anything publishes. Publish less if necessary. One specific, sourced piece outperforms four generic ones.

Practical rules that work:

  1. No brief, no draft. Every piece needs an audience, a single point and evidence before AI writes a word.
  2. Use your own material. Product facts, customer quotes, real examples and first-hand experience.
  3. Source every number. Drop any statistic you cannot trace.
  4. Edit with a kill list. Remove stock phrases and hype on sight.
  5. Review before publishing. A named person asks whether it says something only you could say.

The guide to making AI sound like you covers the setup in more detail.

Skein builds content workflows that start from the business's own facts and customer language, check every claim and stop for approval; the AI engineering page shows how. That suits a team publishing regularly with AI. A business that publishes rarely can get most of the way with a checklist and a careful editor.

Frequently asked questions

What is AI slop?

AI slop is low-quality content made with generative AI: generic, padded, often inaccurate and not requested by the reader. Macquarie Dictionary named "AI slop" its 2025 Word of the Year, defining it as low-quality content created by generative AI, often containing errors and not requested by the user.

How can you tell if content is AI slop?

Check whether a competitor could publish it unchanged, whether it contains specific facts or examples from the business, whether claims are sourced, and whether it relies on stock phrases and padding. If it says nothing only your business could say, it is probably slop.

Is all AI-generated content slop?

No. AI-assisted content can be excellent when it starts from specific material such as real facts, customer insight, evidence and a clear angle, and is edited by a person. Slop comes from asking AI to write with nothing to work from and publishing without review.

Does AI slop hurt SEO?

It tends to perform poorly because it adds nothing new. Search engines and AI assistants favour specific, sourced content, and the 2024 GEO study found citations, statistics and quotations raised visibility in AI answers by up to 40%. Generic content gives them little reason to rank or cite it.

How do you avoid publishing AI slop?

Require a brief with an audience, a single point and evidence before drafting. Feed AI your own facts and customer language, source every number, remove stock phrases, and have a named person review each piece before publishing. Publishing less, but better, usually helps.

Sources
  1. The New Daily. "AI slop is word of the year." The New Daily, 25 November 2025. https://www.thenewdaily.com.au/life/2025/11/25/ai-slop-is-word-of-the-year
  2. Aggarwal, Pranjal, et al. "GEO: Generative Engine Optimization." KDD 2024, arXiv 2311.09735. https://arxiv.org/abs/2311.09735
  3. Chen, Mahe, Xiaoxuan Wang, Kaiwen Chen and Nick Koudas. "Generative Engine Optimization: How to Dominate AI Search." arXiv, 10 September 2025. https://arxiv.org/abs/2509.08919
  4. Pew Research Center. "Google users are less likely to click on links when an AI summary appears in the results." Pew Research Center, 22 July 2025. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/

Talk to us

Find out where AI pays in your business.

Thirty minutes, with the person who would do the work. You leave with the two or three places we would start, whether or not we work together.