Do search engines push AI-written articles down the results, or is the web's range of ideas quietly narrowing instead?

This asks whether search engines like Google push AI-written articles down in results compared with human-written ones. To be direct: nothing in this collection measures Google rankings. No study here compares where AI and human articles land in search results. What the corpus does have is the evidence around that question, and it suggests that 'does AI content rank worse?' may be the wrong frame.

Start with scale. An Internet Archive analysis found that by mid-2025 about 35% of newly published websites were AI-generated or AI-assisted How much of the internet is AI-generated now?. The surprising detail is what didn't change. Factual accuracy and stylistic variety stayed about the same. What dropped was semantic diversity, meaning the range of ideas being expressed, while the overall tone grew more positive. If a ranking system judges pages on surface quality, much of this content would pass. The loss is spread across the whole web instead of showing up on any one page.

The nearest thing to a ranking signal is reader engagement, and there the gap is real but modest. Medium posts that a classifier labeled as AI-written drew about half the likes of human-labeled posts (69 vs. 128) Do readers engage less with AI-generated social media posts?. Disclosing AI help also lowers how readers rate an article, but only slightly, by less than 0.15 points on a 7-point scale. LLM raters apply the same penalty Does disclosing AI assistance make readers trust articles less?. One catch is that people can't actually tell AI writing from human writing: across 30 studies, detection accuracy sits near chance Can people reliably spot content made by AI?. So whatever penalty exists comes from labels and suspicion, not from anything readers can detect. That suspicion can even land on human writers who get wrongly accused of using AI Do unfounded AI accusations harm human writers instead?.

The most useful idea for this question is that search may be the wrong tool. One note contrasts two kinds of information overload. The early internet made existing knowledge much easier to reach, and search and curation solved that. AI content is different because new material keeps being generated with no fixed body of work to rank against. Fixing it takes things like provenance labels and limits on what gets produced, not better ranking Why do search tools fail against AI generated content?. On social platforms, AI posts already win engagement by being thorough, yet they build no lasting reputation for any author. Over time that weakens the platform's job of surfacing trustworthy human voices Does AI content displace human influencers on social media?.

The AI-versus-human split is also blurrier than it sounds. Writers edit AI-drafted paragraphs only 23% of the time, and their edits leave the text about 96% unchanged Do writers actually edit AI-generated text before publishing?. AI assistance also shifts how readers perceive the writer, toward more confident, more extreme, more agreeable voices, across all 29 traits measured Does AI writing assistance change how readers perceive the writer?. Much 'human-written' content in search results is therefore lightly edited AI. So the takeaway isn't whether AI articles rank worse. It's that ranking may no longer be able to tell them apart, and the bigger loss is the shrinking range of ideas.