Author: IRPA AI Senior Analyst, Kieran Gilmurray
Introduction: A Flood of Machine-Written Words
The web is no longer written solely by humans. A recent analysis of 900,000 newly published web pages found that 74.2% contained AI-generated text, meaning that nearly three out of four articles, blogs or product descriptions published today were at least partly machine-written. The internet is being inundated with algorithmic prose at a scale that no editor or search engine can fully track.
This shift has happened almost invisibly. Because AI tools are now built into everyday platforms such as Google Docs, Gmail, LinkedIn and countless CMS systems, using AI to write has become actively difficult to avoid. What started as an occasional productivity hack has become the norm. Of those new pages, only 25.8% were purely human-written, while 15.5% were almost entirely AI-authored.
The same pattern holds in visual media. More than 15 billion AI-generated images were created between 2022 and 2023, roughly one third of all photos ever uploaded to Instagram. People now generate 34 million AI images every day through tools like DALL E, Midjourney and Adobe Firefly. The scale is staggering, and it raises a profound question for readers and viewers alike: in a world of infinite machine-made media, how do we know what is real?
Businesses Embrace Scale, But Lose the Human Touch
Marketers have been among the earliest to embrace generative AI, drawn by its ability to create vast amounts of content at minimal cost. Surveys show that 89% of marketing teams now use generative AI, and 75% say it has become more central to their strategy than the year before. Even small businesses are on board, with two thirds reporting the use of AI tools for content marketing.
But the efficiency comes at a cost. Readers increasingly recognise and resent AI’s generic, formulaic tone. AI text is clean and consistent, but often lacks the warmth or originality that engages people. One NewsGuard analysis of AI-generated news sites found repetitive phrasing, factual errors and even leftover prompts like I cannot complete this request. The result is competent but lifeless content that reads more like a machine report than a human conversation.
Public trust has fallen in tandem. Only 30% of consumers say they trust content they suspect is AI-generated, while more than half feel compelled to verify it. The Cambridge Dictionary even updated its definition of slop in 2025 to mean low-quality content created by AI.
Major publishers have already faced backlash. When the Chicago Sun Times quietly ran a feature article written by AI, readers responded with outrage over the lack of transparency and authenticity. CNET’s ill fated 2023 experiment which used AI to write financial explainers riddled with errors became a case study in reputational damage. The lesson was clear: while AI can scale production, it can also scale distrust.
The Rise of AI Content Farms
A darker consequence of the AI boom has been the return of content farms, this time supercharged by automation. Where once armies of freelancers churned out low value clickbait, a single operator can now run an entire network of AI written sites. NewsGuard identified 49 such sites in mid 2023. By mid 2025, that number had ballooned to over 1,200, spanning 16 languages and posting hundreds of articles a day.
Many of these pages mimic legitimate news outlets, scraping or paraphrasing real reporting to fill their feeds. Some spread falsehoods or outright fabrications, including viral stories that major outlets later had to debunk. The goal is not journalism, it is ad revenue. With negligible production costs, these AI newsrooms flood search results and siphon money away from real publishers.
The damage extends beyond misinformation. As AI-generated pages multiply, they dilute genuine creativity and factual accuracy online. Analysts warn that if AI models continue training on this recycled, low quality content, it could trigger model collapse where future AIs produce even worse results because they are learning from their own flawed output. In effect, the web risks eating itself.
SEO Fallout: Google’s Crackdown and the Humanization Rush
Search engines are pushing back. Google’s 2025 spam policy warns that AI-generated text created solely to boost rankings will be penalized. The company emphasizes Experience, Expertise, Authoritativeness and Trustworthiness regardless of who or what writes the content. Yet in practice, sites stuffed with low value AI copy have seen sharp ranking drops, forcing marketers to rethink their approach. Agencies that once churned out hundreds of AI written posts each week now hire human editors to polish and verify them, while a booming AI humanization industry has emerged to make machine text sound authentic.
Forward looking teams are not abandoning AI but learning to balance it. They use AI for research, outlines and drafts, then rely on people for tone, accuracy and creativity. Adobe’s 2025 Digital Trends report found that 86% of marketing leaders expect AI to speed up production, while 69% plan to invest more in human talent. The future of content is not about choosing between humans and algorithms but blending both to achieve credibility, trust and scale.
The Detection Arms Race
As AI written text dominates the web, an arms race has erupted between content creators and detectors. Companies, universities and publishers have poured millions into algorithms designed to identify machine generated writing, yet none are fully reliable.
Even OpenAI’s own AI Text Classifier was quietly shut down in 2023 after proving accurate only about a quarter of the time. Turnitin’s detector flags roughly 10% of student work as AI written but admits a small false positive rate. Meanwhile, paraphrasing tools like QuillBot and humanizer services are helping users evade detection altogether.
The collateral damage is real. Vanderbilt University found that hundreds of students may have been wrongly accused of using AI after its detector misfired, prompting multiple universities to drop the tools entirely. The market for AI detection is booming, projected to exceed 8 billion dollars by 2032, but its accuracy remains fragile.
Some policymakers argue that the only lasting solution is to build detection into the models themselves. The EU’s AI Act and emerging United States proposals both include requirements for watermarking and provenance verification in generative systems. Until such standards are enforced, however, the cat and mouse game between content creation and detection will continue to escalate.
Implications for Business Leaders
Content volume is no longer a differentiator. AI gives every organisation infinite scale, so value must come from originality, credibility and clear editorial standards.
Trust has become a strategic asset. Readers reward brands that demonstrate transparency and human judgment and penalise those that publish generic AI content.
SEO performance is shifting toward depth and expertise. Low quality AI automation can directly damage rankings and long term visibility.
The risk landscape is expanding. AI driven misinformation, unreliable detectors and synthetic content farms require stronger governance and accountability.
Hybrid content models will lead the market. AI provides speed and efficiency, while humans deliver accuracy, insight, storytelling and voice.
The Way Forward: Restoring Authenticity in an AI-Drenched Web
The AI content surge has forced everyone, from marketers to journalists, to rethink what authenticity means online. Automation can produce limitless words, but it cannot generate human trust.
The most successful organisations are already taking a hybrid approach, pairing AI’s scale with human creativity and oversight. Fact checking, personalisation and transparency are becoming new hallmarks of credible content. Some publishers label AI assisted work, others highlight human reported stories to stand apart.
In the long run, the winners will be those who treat authenticity as a strategy, not sentiment. In a world of infinite machine written text, human insight will be what audiences value most.
About the Author:
Kieran is a globally recognized authority on AI, automation, and digital transformation, having authored multiple influential books and hundreds of articles that have earned him prestigious accolades, including being named a Top 50 Global Thought Leader and Influencer on Generative AI in 2024, a Best LinkedIn Influencer for AI and Marketing, Top 50 Global Thought Leaders and Influencers on Manufacturing 2024, Top 14 people to follow in data and one of the World’s Top 200 Business and Technology Innovators.
CLICK HERE TO SCHEDULE AN ANALYST CHAT
Links:
Originally posted on 2025-12-16 in the IRPA AI Network — Announcements & Updates