Analysts tracking global web request patterns report a 3–5% year-over-year decline in traffic classified as “human-originated” across sampled networks — the first sustained drop observed after years of steady growth.
Chart is illustrative and intended to convey directionality, not precise measurement.
Multiple measurement groups cite a common culprit: automated activity rising faster than human browsing. AI crawlers, search bots, and scraping systems increasingly dominate request volume — with “crawl-to-refer” ratios widening as AI systems ingest content while sending fewer users back to source sites.
“The web’s old bargain — crawl in exchange for referrals — is breaking. Automation is eating the margins.” — Ren Ito, CDN Analyst
| Domain slice (sampled) | Observed pattern | Interpretation |
|---|---|---|
| News & publishing | Automated crawling spikes; referrals soften | AI answer engines reduce click-through demand |
| Reference / how-to | Higher “read without visit” signals | Extraction and summarization displace browsing |
| E-commerce | Stable human traffic, rising bot probes | Price scraping + automation arms race |
Classifying “human-origin” traffic is fuzzy. VPNs, mobile gateways, privacy tooling, and automation that mimics browsers all distort labels. Several experts argue the decline could partly reflect classification drift rather than real behavior change.
“If your classifier improves, the trendline changes. Measurement is politics with better fonts.” — Dr. Lina Mercado, Internet Measurement Researcher