Glossary · noun
AI fatigue
AI fatigue is the tiredness that comes from having to evaluate whether content is authentic, useful or machine-generated before you can use it, and the disengagement that follows when that cost stops seeming worth paying.
In plain English
It is not fatigue caused by AI tools. It is fatigue caused by other people's AI output, and by the extra step it adds to everything you read.
The mechanism
Normally, reading carries an implicit trust that the writer did some work. AI slop breaks that assumption without replacing it, so the reader absorbs a checking cost that used to be paid upstream. Every document gets a second job: work out whether it is worth reading, then read it.
Research framing of the slop problem names this directly. A Columbia University report on AI slop and the information ecosystem lists discernment fatigue among the concerns, alongside saturation and aesthetic degradation: the fear is less about any single bad item than about the cumulative cost of staying oriented.
Why it is rational rather than fragile
There is a version of this that gets framed as a complaint about sensitivity, and that framing is wrong. The behavior is rational. When the cost of verification rises and the expected payoff falls, people check less and skim more, and some of them disengage from channels entirely.
That has consequences that are not about AI at all. A reader who stops trusting a search result, a newsletter or a professional forum does not replace it with a better source, because verifying alternatives carries the same cost.
What it looks like in practice
- Colder, more transactional reading: skim for the specific fact, leave.
- Retreat to a small set of sources that have not yet burned you.
- Abandonment of whole channels, such as a search engine or a retail listing page.
- A version of cynicism in which nothing is checked because everything is suspect.
The last one is the expensive outcome. Generalized suspicion is not discernment; it is the absence of it.
The workplace version
Inside companies the same mechanism shows up as workslop, the AI-generated work content that looks finished and arrives without substance. The 2025 BetterUp and Stanford study of 1,150 US desk workers found 40% had received some in the past month, with each incident costing just under two hours to resolve. That is fatigue with a timesheet.
Origin and history
The phrase follows the phenomenon rather than preceding it. It entered general usage as generated material spread through feeds, search results and workplaces during 2024 and 2025, and it was consolidated by research and reporting that began treating reader exhaustion as a system-level cost rather than a personal complaint.
The related term discernment fatigue appears in the Columbia University report on AI slop and the information ecosystem, alongside aesthetic degradation and saturation, as one of the concerns driving research interest in the topic.
Examples
- Opening a search result and knowing within two sentences that it will not answer the question, four times in a row.
- Scanning a newsletter for one number while skipping everything around it.
- Giving up on a marketplace category because the listings no longer correspond to real products.
- Receiving a document at work and budgeting an hour to work out whether it is usable.
Also known as
- AI exhaustion
- discernment fatigue
- slop fatigue
Sources
Every factual claim on this page is traceable to one of these. Where a number or quotation comes from a single source, it is attributed in the sentence that uses it.
- AI Slop and the Information Ecosystem Columbia IGP report, March 2026
- The Hidden Cost of Workslop (research report) BetterUp Labs and Stanford Social Media Lab
- AI Slop: Definitions and Normative Status MINT Lab
Frequently asked questions
What is AI fatigue?
The tiredness that comes from having to work out whether content is authentic, useful or machine-generated before you can use it, and the disengagement that follows when that cost feels not worth paying.
Is AI fatigue a real thing?
It is not a diagnosis, and it does not need to be one. It describes a rational response to a rise in the cost of verifying what you read, and it is documented as a concern in research on the information environment.
How is AI fatigue different from brain rot?
Brain rot names the effect people attribute to consuming trivial content. AI fatigue is about the effort of sorting good content from bad, which is a cost paid before consumption rather than after.
What helps?
Narrowing sources rather than broadening them, preferring publications that name their sources, and judging artifacts on specificity and accountability rather than trying to detect authorship.