Merriam-Webster just crowned �slop� as 2025�s word of the year�defined as �digital content of low quality� often produced at scale by AI? It�s a punchy cultural signal, but it�s also a hard business truth: the AI content glut is reshaping search traffic, licensing economics, and workplace governance? Are you measuring how much �slop� touches your customers and teams?
Why a dictionary�s choice matters to business
�Slop oozes into everything,� Merriam-Webster writes? Its president, Greg Barlow, called the term �illustrative� of a �transformative technology� that many find �fascinating, annoying, and a little bit ridiculous?� Beneath the wit is a real shift in distribution: AI video generators and summary engines reduce click-throughs to publishers and flood feeds with near-duplicate content that competes on quantity, not quality?
That dynamic shows up in the numbers? A Seer Interactive study found organic click-through rates fell 61% on queries with Google�s AI Overviews, while paid CTRs dropped 68%? Bain & Company estimates AI summaries now appear in at least 40% of searches for 80% of consumers, reducing organic web traffic 15%�25% (ZDNET)? Google disputes some causal claims but concedes not all sites are thriving, while arguing AI �expands� search by enabling longer, more complex queries (ZDNET)?
Licensing vs? links: the fight over who pays for �slop�
Publishers want standardized content licensing for AI training and answers? Google�s search chief Nick Fox said �no� to a universal licensing API, doubling down on sending traffic and links, and highlighting commercial partnerships with 3,000+ publications (ZDNET)? That keeps Google�s model familiar�but leaves many mid-tier outlets exposed to AI answer boxes that satisfy users without a click?
Enter a new idea: �pay-to-crawl?� Creative Commons, historically the champion of open sharing, is now �cautiously supportive� of systems that charge AI bots when they scrape sites�paired with safeguards for researchers and nonprofits (TechCrunch)? Infrastructure firms have begun to align on technical signals like Really Simple Licensing (RSL), now adopted by major CDNs, to express permissions and pricing? The goal: let small and local publishers get paid without negotiating bespoke deals�while avoiding a web that walls itself off?
Inside companies, �slop� becomes a governance risk
The AI glut isn�t just a media problem? At work, AI use is rising and often uncoordinated? Gallup reports 45% of US workers use AI at least a few times a year (23% weekly; 10% daily), while 23% don�t know if their employer has any AI program at all (ZDNET)? That gap invites inconsistent quality, data leakage, and compliance risks? A recent MIT analysis cited by ZDNET found 95% of business AI apps fail to deliver ROI; the standouts pair bottom-up experimentation with explicit training, guardrails, and data access?
The consumer backlash is real, too? Watchdogs caught chatbot-powered toys discussing sexual topics and giving unsafe instructions to children, prompting policy scrutiny and vendor suspensions (Ars Technica)? That�s �slop� in its most sensitive context�and a preview of reputational risk when guardrails fail?
Regulation is coming�just not evenly
Policy is amplifying the uncertainty? A recent executive order seeks to preempt some state AI rules and push toward a single federal framework, directing federal challenges to state laws while Congress hashes out standards (TechCrunch; Ars Technica)? Supporters say patchwork rules stifle startups; critics warn it weakens consumer protections? Either way, the rulebook is in flux, and legal limbo raises the cost of poorly governed �slop?�
What leaders can do this quarter
- Define �quality bars� by workflow: publish simple, auditable criteria for acceptable AI outputs in marketing, code, and ops?
- Instrument AI usage: track prompts, outputs, and destinations; log what is externally published to reduce public-facing �slop?�
- Harden data paths: whitelist tools, add safety training, and restrict sensitive data in prompts; adopt retrieval and policy checks?
- Plan for licensing signals: monitor RSL/pay-to-crawl support in your stack and prepare to respect (and benefit from) machine-readable terms?
Word of the year or not, �slop� is now a business KPI? Companies that treat it as an exposure to be measured�and a standard to be raised�will fare better in 2026 than those that wait for the next algorithm change to decide for them?