State of the Automated Internet
How much of the internet is machines? A quarterly, fully sourced snapshot: bot traffic share, AI-generated content…
Every claim on this site is supposed to carry a source. This page is where the sources live. Four times a year the writer gathers the newest published measurements of how much of the internet is machine-made — traffic, text, models, detection, fraud — puts them in one place, dates them, and notes what they do not prove. Numbers below are the most recent published figures as of July 2026.
① Traffic: the machines hold the majority
The thirteenth annual Bad Bot Report, published April 2026 by Imperva (a Thales company), found that automated traffic accounted for more than 53% of all web traffic in 2025, up from 51% the year before. Human activity fell to roughly 47%. This was the second consecutive year machines outnumbered people, which matters more than the first: one crossing is an anomaly, two is a trend line.Source: Imperva / Thales, 2026 Bad Bot Report, published Apr 2026, based on full-year 2025 data.
The same report attributes 40% of all web traffic to malicious bots specifically, a three-point rise on the previous year. The traffic is not only inhuman; increasingly it does not even pass through the human-facing internet.
"Bots" is not a synonym for "fakes." Roughly a quarter of that automated share is legitimate: search engine crawlers, uptime monitors, accessibility tools, RSS readers, and the AI crawlers that index the web — including the ones that may have brought you here. The dead-internet claim is about who the content is for, not merely who requests it.
② Text: about half of new articles, and holding
Research firm Graphite sampled English-language articles from Common Crawl and classified them with multiple detection tools. Their finding, reported by Axios in May 2026: primarily AI-generated articles reached 35.9% within a year of ChatGPT's launch, about 48% by late 2024, and have hovered near half ever since — an estimated ~50% in Q1 2026.Source: Graphite analysis, reported by Axios and Search Engine Land, May 2026.
The widely repeated prediction that 90% of online content would be synthetic by 2026 — traceable to a 2022 Europol report — did not happen. The curve flattened. Publishing volume is not the same as consumption: Graphite also found that AI-generated articles are published in enormous quantity but appear far less often among top-ranking search results, which still skew human-written or heavily human-edited.
Nobody can cleanly count this. Most writing now passes through a human and a model — outlined by one, drafted by the other, edited back. Detectors classify a spectrum as a binary. Read every "X% of the web is AI" figure, including the ones on this page, as an estimate produced by imperfect detectors, not a census.
③ Detection: signals, not proof
The tools that produce those percentages are weaker than their marketing. The RAID benchmark (University of Pennsylvania, ACL 2024) evaluated twelve detectors against 6.2 million generations across eleven models, eight domains and eleven adversarial attacks, finding substantial difficulty generalising to unseen models and domains and steep accuracy drops from simple changes such as synonym swaps or a repetition penalty. Stanford research published in Patterns in July 2023 tested seven widely used detectors on 91 TOEFL essays written by non-native English speakers and found an average false-positive rate of 61.2%, while essays by native writers were classified with near-perfect accuracy.Sources: Dugan et al., “RAID: A Shared Benchmark for Robust Evaluation of Machine-Generated Text Detectors”, ACL 2024 (6.2m generations, 11 models, 8 domains, 11 adversarial attacks; 12 detectors evaluated) · Liang, Yuksekgonul, Mao, Wu & Zou, “GPT detectors are biased against non-native English writers”, Patterns 4(7):100779, July 2023.
Vendors publish accuracy figures of 98–99%, measured on raw, unedited model output against clean human text. Independent testing on realistic material — edited, paraphrased, mixed — lands far lower. This is why the museum's Inspector stamps every finding SIGNALS, NOT PROOF, and why no one should face consequences on a detector score alone.
④ Models: the shipping rate
Public trackers count 198 frontier models shipped since ChatGPT's launch in November 2022. In mid-2026 the frontier tier itself widened from three labs to five, with a wave of open-weights releases one stratum below.Source: aireleasetracker.com and lab announcements, recorded Jul 2026. See THE MOVERS.
⑤ Fraud: the first year AI got its own line
In April 2026 the FBI published its 2025 Internet Crime Report. For the first time in the Internet Crime Complaint Center's 25-year history, artificial intelligence was broken out as a tracked category: 22,364 complaints and $893,346,472 in reported losses. Investment fraud accounted for roughly $632M of it; business email compromise with a confirmed AI component more than $30M; deepfake job-interview schemes about $13M.Source: FBI, 2025 Internet Crime Report (IC3), published Apr 2026.
Total reported cybercrime losses reached $20.877 billion across 1,008,597 complaints, and Americans over 60 reported roughly $7.7 billion in fraud losses — the steepest rise of any age group. The FBI itself notes the AI figure is almost certainly an undercount: it counts only cases where the victim recognized that AI was involved. A cloned voice that worked leaves no evidence that it was cloned.
The five numbers, in one table
| Measure | Figure | Period | Source (date published) | Direction |
|---|---|---|---|---|
| Automated share of web traffic | 53% | Full-year 2025 | Imperva / Thales Bad Bot Report (Apr 2026) | Rising |
| New articles primarily AI-generated | ~50% | Q1 2026 | Graphite, via Axios (May 2026) | Flat since early 2025 |
| AI-related fraud losses (US, reported) | $893M | Full-year 2025 | FBI IC3 Internet Crime Report (Apr 2026) | New category |
| Frontier models shipped since ChatGPT | 198 | Nov 2022 – Jul 2026 | Public release trackers (Jul 2026) | Rising |
| Detector false-positive rate, non-native English writers | >60% | 2023 study | Liang et al., Patterns (2023) | Unresolved |
Method, and what this page is not
The full sourcing rules live on the method page; every correction is logged publicly on the corrections page.
- Every figure comes from a named, published source with a date printed beside it. Where a number is contested, the dispute is stated.
- Primary sources are preferred: the report or paper itself, or first-hand reporting of it. Vendor marketing claims and SEO roundups are excluded.
- Nothing here is original measurement. the writer does not crawl the web; this page gathers, dates, and contextualizes the work of others.
- Figures are snapshots. Sources publish on their own schedules — annually, continuously, or irregularly — and this page is re-cut every quarter around whatever is newest.
- Superseded numbers move to the archive below rather than disappearing, so the trend line stays visible.
The measurements are real and the direction is uncomfortable — but the honest reading of this quarter is not that the internet ended. Automated traffic passed half; a large share of it is ordinary infrastructure. AI writing reached about half of new articles and then stopped climbing, and it still loses to human work where readers actually are. Detection remains too weak to accuse anyone. The genuine harm with a hard number attached is fraud, and it lands hardest on the elderly. Worry is warranted. Hostility toward the technology is not the same thing as attention to the evidence.
Archive
Q3 2026 — this edition. First publication. Headline figures: bots 53% of traffic (2025), AI-written articles ~50% (Q1 2026), AI fraud losses $893M (2025), 198 frontier models since Nov 2022.
Q4 2026 — due October 2026.
Sources
- Imperva / Thales — 2026 Bad Bot Report: Bots in the Agentic Age, published April 2026 (full-year 2025 data).
- Graphite — analysis of AI-generated article share, reported by Axios and Search Engine Land, May 2026.
- Ahrefs — study of AI content in new web pages, April 2025.
- FBI Internet Crime Complaint Center — 2025 Internet Crime Report, published April 2026.
- RAID benchmark — University of Pennsylvania, ACL 2024.
- Liang et al. — GPT detectors are biased against non-native English writers, Patterns, 2023.
- University of Chicago Booth — AI detector false-positive study, August 2025.
- Public model release trackers and lab announcements, recorded July 2026.