AI Slop

AI slop is digital content created with generative artificial intelligence that is perceived as lacking effort, quality or meaning, and produced in high volume — usually to gain attention, earn advertising revenue, or deceive. The term carries a pejorative connotation comparable to spam, and in late 2025 it was selected as the Word of the Year by both Merriam-Webster and the American Dialect Society.^[1][2]

Background

The word slop entered English around 1400 from Old English roots meaning dung and slime, later denoting animal feed made from food waste before generalising to "rubbish". Early uses of slop to describe low-grade AI material emerged in online communities — 4chan, Hacker News and YouTube — apparently in reaction to AI image generators released from 2022 onward. British programmer Simon Willison is credited with championing the term in the mainstream, writing about it on his personal blog in May 2024 while noting it predated his advocacy.^[1]

Competing terms — "AI garbage", "AI pollution", "AI-generated dross" — were proposed but failed to stick. Usage surged in the second quarter of 2024, driven partly by Google's AI-generated search answers, and media criticism of the flood of low-quality AI content peaked through the fourth quarter of 2024. The Cambridge Dictionary formally added an AI-related definition of slop on 24 June 2025: "content on the internet that is of very low quality, especially when it is created by artificial intelligence".^[3]

Key Concepts

Academic work has struggled to fix a precise boundary. A January 2026 paper by Cody Kommers and colleagues argued AI slop has "so far resisted formal definition" while identifying three prototypical properties:^[4]

• Superficial competence — the output looks plausible and well-formed at a glance.
• Asymmetric effort — producing it costs far less than the effort an equivalent human-made item would have required.
• Mass producibility — it can be generated at effectively unlimited scale.

The economic logic was captured by Oliver Whang in The New York Times: "The goal of pig slop is to maximize nutrient intake while minimizing cost; the goal of A.I. slop is to maximize time spent consuming content while minimizing cost."^[1] Former Evernote CEO Phil Libin drew the intent-based line: AI used to make mediocre things with less effort than they'd otherwise take is slop; AI used to make something better than possible without it is positive augmentation.^[2]

Derivative coinages followed: workslop (Harvard Business Review and Stanford/BetterUp research found 40% of surveyed employees received AI-generated "workslop" that looked good but lacked substance, each incident costing colleagues about two hours to resolve), trendslop, hiring slop for AI-flooded résumés, and slopper for a user overly reliant on generative tools.^[1][2]

Why It Matters for AI

AI slop is the visible face of generative AI's cost collapse meeting the attention economy. It pollutes search results and social feeds, erodes trust in authentic media, and is exploited in political campaigning and scams. Slop also runs in both directions: a Harvard Business Review study described AI-generated résumés degrading hiring pipelines, and AI-generated music charted — an AI song topped Billboard's country digital song sales chart in November 2025. The phenomenon has made disclosure, provenance and content-authenticity tooling (see AI watermarking) a mainstream policy concern.^[2]

>Key Takeaways

• AI slop is bulk-produced generative-AI content perceived as low-effort or meaningless, made for attention or profit.^[1]
• The term moved from online slang to dictionaries in 2024–25; Merriam-Webster and the American Dialect Society named it 2025 Word of the Year.^[2]
• Researchers describe it via three properties: superficial competence, asymmetric effort and mass producibility.^[4]
• It spawns workplace and civic variants — workslop, hiring slop, political content farming — each with measurable costs.
• For Malaysia, the practical responses are media literacy, content provenance/labelling, and human-review workflows in AI-assisted work.

See Also

• AI Content Detection
• AI Watermarking
• AI Content Moderation
• AI Ethics

References