# AEO Content: Free AI Content Detector ## Cheap AI content is easy to spot. Find out how yours reads. Paste text, a page address or a file. See who wrote it, how it reads, and whether AI answers like ChatGPT will cite it. ## One tool, three readings ### AI Content Detector Three readings of any text or page: who wrote it, how it reads, and how ready it is to be cited in AI answers. Paste the text, drop in a page address, or upload a file. 1. Paste text, a page address or a file - Up to 30,000 characters, any public page, or a .txt, .md or .csv file. Free, no sign-up, and nothing you paste is stored. 2. Read who wrote it - A trained model makes one of three calls: written by AI, written by a person, or not called. It calls AI only when a passage is unmistakably machine-written. 3. Read how it reads - A deterministic score out of 100 from seven weighted measurements of the writing, where higher is better writing: Distinctive from 70, Average from 45, Generic below. The phrases that read as machine-written are listed, and marked in pasted text. 4. Read AI answers - AEO Rank, the engine behind our site audits, run on this one text: a score out of 100, where 70 is the mark where an answer engine has something it can lift and cite. Ready from 70, partly ready from 40, not ready below. The fixes that matter most are listed. 5. Send the result - One link carries the score, the verdict and the full breakdown. ## Common questions ### Does this tell me whether AI wrote the text? It gives two separate readings and never blends them. How it reads is a deterministic measurement of the writing - sentence rhythm, point of view, vocabulary, stock phrases. Who wrote it is a trained model that makes one of three calls: written by AI, written by a person, or not called. It says AI only when a passage is unmistakably machine-written; when nothing is, it says so instead of guessing. A person can write generic copy and a good AI can write with voice - both readings can be true at once. ### How does the "who wrote it" call work, and how sure is it? The text is cut into passages of about 300 words and each one is read by a classifier we trained on our own AI output and on writing we know people wrote. The document is judged by its most machine-like passage, because long AI drafts often have a machine-written opening and a body that reads human. The bar for an AI call is set high on purpose - on thousands of documents we know people wrote, almost none reached it - and there is a second model behind it that reports how predictable the writing is and which sentences pull the reading either way. Treat "not called" as exactly that: inconclusive, not a soft accusation. Paraphrased AI can still land there. ### What does the score actually mean? Higher is better writing. Seven measurements of the prose itself are weighted into one number: sentence rhythm counts most at 22 percent, stock phrases 20, point of view 14, short-sentence mix and active voice 12 each, facts versus interpretation and vocabulary range 10 each. Open the breakdown in any result to see all seven. The score says nothing about who wrote the text - that is the other reading. ### Is it deterministic? The style score is: no model runs in it, and the same text always returns the same number, which is what makes a re-check after an edit worth anything. The who-wrote-it call comes from a model and is stable for the same text, but it is a call, not a measurement - which is why it is shown as words, never as a percentage. ### How much text do I need? At least 150 words for a reading you can rely on. Below that the ratios swing on a single sentence, and the result says so rather than pretending otherwise. The who-wrote-it call needs at least one full passage of about 300 words to be worth much; on a short text it will more often say "not called". ### What languages does it handle? English only. Every word list and pattern in the style engine is English, and the origin model was trained on English, so a text in another language gets flagged instead of scored. ### Do you keep what I paste? No. The text is sent to the scoring endpoint and to the model that reads it, measured, and never written anywhere. We keep only counts - how many checks, what the calls were, how long they took - never the text. ## Research How AI writes and how people write. Index: https://aeocontent.com/research/ ### How AI Writes vs How Humans Write: What Nearly 10,000 Verified Human Documents Showed Us URL: https://aeocontent.com/research/how-ai-writes-vs-how-humans-write/ Role: pillar article of the research cluster Author: Alex Shortov, CTO, AEO Content Published: 2026-08-31; last modified: 2026-08-31 We tested an AI detector on 9,963 verified human documents and hundreds of AI-drafted articles. What separates AI and human writing once surface tells are gone. AI writing and human writing differ in origin, not in quality. Surface tells like dashes and stock phrases can be edited out. What stays is the model's underlying style and the absence of first-hand specifics. ### AI Detector False Positives: What a 0.5% Cap Really Means URL: https://aeocontent.com/research/ai-detector-false-positives-what-a-0-5-cap-really-means/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-15; last modified: 2026-09-15 Why AI detector false positive rates vary by writing register, what the 0.5% per-register cap means, and how the not-called zone protects your work. An AI detector false-positive rate is the share of verified human documents a detector incorrectly calls AI. That rate can vary by type of writing: our previous detector called 0.14% of pre-2021 Medium posts AI but 1.26% of US federal agency documents. ### How AI Detectors Work and Where They Break URL: https://aeocontent.com/research/how-ai-detectors-work-and-where-they-break/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-14; last modified: 2026-09-14 AI detectors split into classifiers and zero-shot methods with distinct failure modes. First-hand rebuild data covers calibration, paraphrase resistance, and institutional policy. AI detection refers to tools that estimate whether a text was written by a language model or by a person. Two broad families exist. Trained classifiers, such as the open desklib and Fakespot detectors, learn from labeled examples. ### How to Prove a Text Was Written by a Human URL: https://aeocontent.com/research/how-to-prove-a-text-was-written-by-a-human/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-11; last modified: 2026-09-11 Only a Wayback Machine or Common Crawl snapshot before Dec 31, 2020 proves human authorship. Learn the 3-tier evidence framework from AEO Content. Proving that a text was written by a human means establishing provenance , not judging style. The strongest evidence we use is an independent archive capture: a Wayback Machine or Common Crawl snapshot showing the words existed at a known date. ### How One Company Blog Went From Human Writers to AI, Year by Year URL: https://aeocontent.com/research/how-one-company-blog-went-from-human-writers-to-ai-year-by-year/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-10; last modified: 2026-09-10 502 dated posts from one payments blog scored era by era show flagged-AI content reaching 100% after polishing. Bylines are unreliable. Origin detection holds when word filters fail. We traced how one company blog moved from human writers to AI, using 502 dated posts that a payments company published between August 2018 and September 2026, and compared it with a sports equipment manufacturer's site. ### What Makes AI-Drafted Content Read Human: First-Hand Data URL: https://aeocontent.com/research/what-makes-ai-drafted-content-read-human-first-hand-data/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-09; last modified: 2026-09-09 First-party data from 9,963 verified documents shows why proprietary records make AI content read human, while humanizer tools and synonym swaps do not. Give the draft material only your company could publish, then edit the surface. On one site we measured, record-based sections read AI 37% of the time, against 71% for the rest. ### Why Banning AI Buzzwords Does Not Make AI Text Read Human URL: https://aeocontent.com/research/why-banning-ai-buzzwords-does-not-make-ai-text-read-human/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-08; last modified: 2026-09-08 Phrase bans on 816 AI articles hit 98-100% compliance. Detection held at 81-100%. The substance gap is what matters, and word lists can't close it. AI words-to-avoid lists are collections of phrases that appear more often in AI output than in human writing, which editors remove so drafts stop sounding templated. ### The Em Dash Is Not the AI Tell. The Dash Rate Is. URL: https://aeocontent.com/research/the-em-dash-is-not-the-ai-tell-the-dash-rate-is/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-04; last modified: 2026-09-04 The em dash isn't the AI writing tell. The dash rate is. What research across 12 AI models and verified human documents shows about detection signals. Not on its own. A high rate of dash-punctuated asides per 1,000 words is the stronger tell, whichever dash character carries them, and even that rate is only one signal. ### Why Human Writing Gets Flagged as AI, and Which Writing Gets Flagged Most URL: https://aeocontent.com/research/why-human-writing-gets-flagged-as-ai-and-which-writing-gets-flagged-most/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-03; last modified: 2026-09-03 Why formal prose is flagged as AI: it shares AI's default register. Which writing types trigger most false positives and what calibration does about it. An AI detection false positive is a human-written document that a detector calls AI. In our tests, the human writing misread most often was neutral institutional prose: a government program update, a policy analysis, a corporate case study. ### Which Parts of an AI-Written Article Give It Away URL: https://aeocontent.com/research/which-parts-of-an-ai-written-article-give-it-away/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-02; last modified: 2026-09-02 Structural patterns reveal AI writing, not words. Detection rates from 2,625 passages: decision frameworks 100%, author bios 8%. First-party September 2026 data. An AI-written article gives itself away through the shape of its sections more than through individual words. ### Do Humanized AI Articles Pass AI Detectors? We Tested Four URL: https://aeocontent.com/research/do-humanized-ai-articles-pass-ai-detectors-we-tested-four/ Author: Alex Shortov, CTO, AEO Content Published: 2026-09-01; last modified: 2026-09-01 223 pipeline articles, four calibrated detectors, zero flags at strict thresholds. AUROC scores, methodology, and what determines whether AI content passes detection. Against general-purpose detectors, well-edited AI articles can pass: four open detectors flagged 0 of 223 held-out AEO Content articles at a 1% human false-positive threshold. ## News What changed in AI writing and AI search. Index: https://aeocontent.com/news/ ### New AI Detection Technology Is Changing How News Websites Create and Publish Content URL: https://aeocontent.com/news/new-ai-detection-technology-is-changing-how-news-websites-3b3e24/ Original story: https://goodmenproject.com/technology/new-ai-detection-technology-is-changing-how-news-websites-create-and-publish-content/ (The Good Men Project, 2026-09-15) Our note published: 2026-09-15 AI detection tools are changing editorial gatekeeping at news sites, with direct consequences for writers who use AI assistance. ### Google Tests Paying Publishers For AI Answers Via Search Console URL: https://aeocontent.com/news/google-tests-paying-publishers-for-ai-answers-via-search-a9ce22/ Original story: https://news.google.com/rss/articles/CBMirAFBVV95cUxPYWNQVkRVZTNfbTRjQXQ1UWk1ZWFHZVVVaE84dHFNRHpXc0NqQVcwS3htbEZJUWVjV2tlbkR1cUwyUW9yM3BTLXh3QVBNRGgyQ0VpSUdYMTgtRld2bFlrT2hDc011eDNyT0lwcVdzdnhWMTRVT2FESWVxbVd3ZDdTbmJTUm43S2JfcWc4NVVKX21xbHRNaFJCSmZnRE1wZVFDYUg2M2dvM2t3dVBP?oc=5 (Search Engine Journal, 2026-09-14) Our note published: 2026-09-15 Google testing payments to publishers whose content feeds AI answers could shift how writers and outlets think about producing AI-readable material. ### What Wikipedia Reveals About AI Overviews And Web Traffic URL: https://aeocontent.com/news/what-wikipedia-reveals-about-ai-overviews-and-web-traffic-39d6e8/ Original story: https://www.searchenginejournal.com/what-wikipedia-reveals-about-ai-overviews-and-web-traffic/589042/ (Search Engine Journal, 2026-09-11) Our note published: 2026-09-15 Wikipedia's traffic patterns offer concrete data on how AI Overviews select and deprioritize web sources, with implications for any publisher seeking citations. ### Is Your Favourite Political Influencer Real? 'Ghost Creators' Are Paid to Read AI Scripts Attacking Democrats URL: https://aeocontent.com/news/is-your-favourite-political-influencer-real-ghost-creators-are-d49bbb/ Original story: https://www.ibtimes.co.uk/ai-driven-political-content-ghost-creators-youtube-1818509 (International Business Times UK, 2026-09-08) Our note published: 2026-09-15 Political operatives paid to perform AI-written scripts while posing as authentic influencers shows how synthetic content can evade detection in high-stakes public discourse. ### AI Slop Cleanup Job Listings Up 87%, Report Says URL: https://aeocontent.com/news/ai-slop-cleanup-job-listings-up-87-report-says-0ad926/ Original story: https://www.searchenginejournal.com/ai-slop-cleanup-job-listings-report/588235/ (Search Engine Journal, 2026-09-03) Our note published: 2026-09-15 An 87% jump in job listings for AI content cleanup shows how widely low-quality AI-generated writing has spread and how publishers are now paying to fix it. ### Can Colleges Detect ChatGPT? Detection Methods 2026 URL: https://aeocontent.com/news/can-colleges-detect-chatgpt-detection-methods-2026-3ad70e/ Original story: https://undetectable.ai/blog/can-colleges-detect-chatgpt/ (Undetectable AI, 2026-08-30) Our note published: 2026-09-15 A 2026 rundown of how colleges try to catch AI writing — useful as a field overview, but the source is a commercial humanizer with a clear stake in the findings. ### Creator-First Answer Engine Optimization Takes the Stage URL: https://aeocontent.com/news/creator-first-answer-engine-optimization-takes-the-stage-da5a25/ Original story: https://www.wsj.com/cmo-today/creator-first-answer-engine-optimization-takes-the-stage-04b28060 (WSJ, 2026-08-26) Our note published: 2026-09-15 WSJ coverage of creator-led AEO signals that writing to be cited by AI answer engines is entering mainstream content strategy. ### Reports Indicate Google’s Spam Update Focused On SEO AI Content URL: https://aeocontent.com/news/reports-indicate-googles-spam-update-focused-on-seo-ai-6af598/ Original story: https://www.searchenginejournal.com/reports-indicate-googles-spam-update-focused-on-seo-ai-content/586978/ (Search Engine Journal, 2026-08-25) Our note published: 2026-09-15 Google's latest spam update appears to have specifically penalised AI-generated SEO content, signalling stricter algorithmic enforcement against AI writing at scale. ### Can AI Detectors be Wrong? (Learn How To Avoid AI Detection) URL: https://aeocontent.com/news/can-ai-detectors-be-wrong-learn-how-to-avoid-da8a35/ Original story: https://undetectable.ai/blog/can-ai-detectors-be-wrong/ (Undetectable AI, 2026-08-21) Our note published: 2026-09-15 Undetectable AI makes the case that current detectors produce significant false positives, framing accuracy gaps as justification for its own humanising tool. ### Why Reddit’s ChatGPT Citation Drop Isn’t Fully Explained URL: https://aeocontent.com/news/why-reddits-chatgpt-citation-drop-isnt-fully-explained-27aede/ Original story: https://www.searchenginejournal.com/why-reddits-chatgpt-citation-drop-isnt-fully-explained/586479/ (Search Engine Journal, 2026-08-19) Our note published: 2026-09-15 ChatGPT is citing Reddit far less than it used to, but analysts cannot agree on whether a model change, a licensing dispute, or a crawl decision is responsible — a gap that matters for anyone tracking how AI picks its sources. ### How Google AIOs Use Facebook, Instagram & TikTok From 300 Million US Searches URL: https://aeocontent.com/news/how-google-aios-use-facebook-instagram-tiktok-from-300-131f56/ Original story: https://www.searchenginejournal.com/how-google-aios-use-facebook-instagram-tiktok-from-300-million-us-searches/585244/ (Search Engine Journal, 2026-08-18) Our note published: 2026-09-15 Large-scale data shows Google AI Overviews routinely surface social-platform content, reshaping which sources matter for AI-driven visibility. ### Half of Every HSC Grade Is Unenforceable: NSW Ends Take-Home Work Over AI Detection Failure URL: https://aeocontent.com/news/half-of-every-hsc-grade-is-unenforceable-nsw-ends-82b0d7/ Original story: https://www.techtimes.com/articles/323700/20260810/half-every-hsc-grade-unenforceable-nsw-ends-take-home-work-over-ai-detection-failure.htm (Tech Times, 2026-08-10) Our note published: 2026-09-15 New South Wales is eliminating take-home work after AI detectors repeatedly failed to produce fair, reliable results — an institutional admission that current tools cannot police AI writing in schools. ### What Top Stories Inside AI Overviews Means For Publishers And Brands In 2026 And Beyond URL: https://aeocontent.com/news/what-top-stories-inside-ai-overviews-means-for-publishers-9de15c/ Original story: https://www.searchenginejournal.com/what-top-stories-inside-ai-overviews-means-for-publishers-and-brands-in-2026-and-beyond/584175/ (Search Engine Journal, 2026-08-04) Our note published: 2026-09-15 Google is inserting Top Stories carousels inside AI Overviews, and the article maps out which publisher content wins that new real estate in 2026. ### California's AI Transparency Act Takes Effect With Fines That Compound Daily URL: https://aeocontent.com/news/californias-ai-transparency-act-takes-effect-with-fines-that-f32d63/ Original story: https://startupfortune.com/californias-ai-transparency-act-takes-effect-with-fines-that-compound-daily/ (Startup Fortune, 2026-08-01) Our note published: 2026-09-15 California's AI Transparency Act is now live with compounding daily fines, creating real legal exposure for any publisher that fails to label AI-generated content. ### Google Search Results Still Rank Heavily AI-Flagged Pages URL: https://aeocontent.com/news/google-search-results-still-rank-heavily-ai-flagged-pages-3bfa9b/ Original story: https://www.searchenginejournal.com/heavily-ai-flagged-pages-still-rank-across-googles-top-10/583983/ (Search Engine Journal, 2026-07-28) Our note published: 2026-09-15 Google continuing to rank pages flagged as AI-heavy signals that detection scores alone do not trigger a ranking penalty, a key data point for anyone advising on AI content strategy. ### Google May Be Penalizing AI-Generated Content As Thin Content URL: https://aeocontent.com/news/google-may-be-penalizing-ai-generated-content-as-thin-d211eb/ Original story: https://www.searchenginejournal.com/google-may-be-penalizing-ai-generated-content-as-thin-content/583773/ (Search Engine Journal, 2026-07-27) Our note published: 2026-09-15 Google appears to be penalizing AI-generated content as thin content, a direct warning for anyone publishing AI-written text at scale on their website. ### YouTube Cracks Down on Mass-Produced AI Content, Cutting Off Ad Revenue for Low-Quality 'Slop' Channels URL: https://aeocontent.com/news/youtube-cracks-down-on-mass-produced-ai-content-cutting-2c9f73/ Original story: https://www.ibtimes.com.au/youtube-tightens-rules-ai-generated-content-monetization-1872928 (International Business Times Australia, 2026-07-26) Our note published: 2026-09-15 YouTube cutting ad revenue from mass-produced AI channels is the most direct platform enforcement yet against low-quality AI content farms, setting a precedent other platforms may follow. ### The Vatican entered into a dispute with AI detectors over the first encyclical of Leo XIV - hi-Tech.ua URL: https://aeocontent.com/news/the-vatican-entered-into-a-dispute-with-ai-detectors-802376/ Original story: https://hi-tech.ua/en/the-vatican-entered-into-a-dispute-with-ai-detectors-over-the-first-encyclical-of-leo-xiv/ (hi-Tech.ua, 2026-07-24) Our note published: 2026-09-15 AI detectors flagging an official papal document as machine-written is a high-profile illustration of the false-positive risk that undermines trust in these tools. ### Substack's new AI scanner tells paying readers exactly how much of their newsletter a human actually wrote URL: https://aeocontent.com/news/substacks-new-ai-scanner-tells-paying-readers-exactly-how-753a34/ Original story: https://startupfortune.com/substacks-new-ai-scanner-tells-paying-readers-exactly-how-much-of-their-newsletter-a-human-actually-wrote/ (Startup Fortune, 2026-07-22) Our note published: 2026-09-15 Substack's new per-newsletter AI meter gives paying readers a transparency signal on human vs. machine authorship, a move that could pressure other publishing platforms to follow. ### AI or Not Detects 98% of Cropped Meta AI Images. Meta's Own Tool Catches 35%. URL: https://aeocontent.com/news/ai-or-not-detects-98-of-cropped-meta-ai-0d0f96/ Original story: https://www.aiornot.com/blog/ai-detection/meta-ai-cropped-image-detection (aiornot.com, 2026-07-19) Our note published: 2026-09-15 A 63-point accuracy gap between a specialist detector and Meta's own provenance tool reveals how far behind platform-native content authentication still lags. ### AI Visibility Rankings Aren’t Stable – New Research Shows It’s Mostly Statistical Noise URL: https://aeocontent.com/news/ai-visibility-rankings-arent-stable-new-research-shows-its-d4572c/ Original story: https://www.searchenginejournal.com/ai-visibility-rankings-arent-stable-new-research-shows-its-mostly-statistical-noise/581905/ (Search Engine Journal, 2026-07-11) Our note published: 2026-09-15 New research finds that brands' positions in AI-generated answers shift too randomly to be gamed, making consistent content quality more important than visibility tactics. ### Google just wiped out 50,000 AI slop networks. Marketers, pay attention URL: https://aeocontent.com/news/google-just-wiped-out-50-000-ai-slop-networks-3ee4f2/ Original story: https://www.thedrum.com/industry-insight/google-just-wiped-out-50-000-ai-slop-networks-marketers-pay-attention (The Drum, 2026-07-08) Our note published: 2026-09-15 Google's mass deindexing of 50,000 AI-generated spam networks is the largest enforcement action yet against AI content manipulation—a clear escalation in how the search giant polices AI writing at scale. ### Google Data Shows AI Search Users Moved Past Keywords, Your Content Hasn’t URL: https://aeocontent.com/news/google-data-shows-ai-search-users-moved-past-keywords-062f72/ Original story: https://www.searchenginejournal.com/google-data-shows-ai-search-users-moved-past-keywords-your-content-hasnt/580596/ (Search Engine Journal, 2026-07-02) Our note published: 2026-09-15 Google's own data confirms AI search users ask in plain language rather than keywords, but most publisher content hasn't adapted to match. ### Google AI Overviews Study Finds Lost Clicks Weren’t Lower Quality URL: https://aeocontent.com/news/google-ai-overviews-study-finds-lost-clicks-werent-lower-fd0a9e/ Original story: https://www.searchenginejournal.com/google-ai-overviews-study-finds-lost-clicks-werent-lower-quality/581298/ (Search Engine Journal, 2026-07-01) Our note published: 2026-09-15 A study finds clicks displaced by Google AI Overviews were not low-quality, suggesting AI summaries divert engaged readers rather than casual browsers. ### Anthropic’s Claude Fable 5 Is Back With New Usage Limits And Safeguards URL: https://aeocontent.com/news/anthropics-claude-fable-5-is-back-with-new-usage-5d5820/ Original story: https://www.searchenginejournal.com/anthropics-claude-fable-5-is-back-with-new-usage-limits-and-safeguards/581231/ (Search Engine Journal, 2026-07-01) Our note published: 2026-09-15 A returning Claude model ships with revised usage caps and new safeguards, which will shape how the model generates and controls written output. ## Machine-readable files - LLM manifest: https://aeocontent.com/llms.txt - RSS feed: https://aeocontent.com/rss.xml - AI usage policy: https://aeocontent.com/ai.txt - Sitemap: https://aeocontent.com/sitemap-index.xml - robots.txt: https://aeocontent.com/robots.txt ## Related sites - AEO Content platform: https://www.aeocontent.ai - Free AI visibility audit: https://audit.aeocontent.ai ## Contact Email: hello@aeocontent.ai Operator: AEO Content, Inc.