Rate: 2.454 /secWorldwide. Derived estimate: ~4% of an estimated 5.3 billion photos taken per day (Photutorial 2026; DataReportal Digital 2026 smartphone base) ≈ 212.0 million selfies selfies per day. Converted to selfies per second.
Rate: 318.287 /secWorldwide. About 4.5 hours per day of online video per internet user × ~6.12 billion users = 27.5 billion hours hours per day. Converted to hours per second.
Rate: 200 /dayWorldwide. UK alone: 12,000+ arrests in 2023 (~33 per day). Global estimate from UK, Belarus, Germany, China, Turkey, Russia and others = 200+ per day. Counter shows estimated arrests today.
Rate: 65 /secWorldwide. WHO HBSC 2024: ~15% of adolescents cyberbullied in the past couple of months (44 countries, 279,000 respondents). Triangulated with UN SRSG 2026 (~1 in 5 globally) and UNESCO 2024 (~1 in 10). Modelled daily estimate: prevalence ÷ 30 days × 1.0B online youth ≈ 5.6 million victims/day. Note: this is a modelled rate from prevalence surveys, not a directly measured daily count.
Rate: 4 /secWorldwide. FBI IC3 2025: 1 million+ US complaints; only ~10-15% of victims report. Conservative global estimate: ~110 million victims/year ÷ 31,557,600 seconds ≈ 3.5 victims/sec. Note: GASA/Feedzai report ~50% of consumers encounter scam attempts weekly, a far broader measure than confirmed victims.
Rate: 70 /hrWorldwide. No single global count. Scaled from US FBI reported victims and FTC data; FTC reported 55,604 romance scam victims in Jan-Sep 2025 (+22% vs same period 2024). US assumed to represent ~20-25% of global victims; converted to average per hour.
Rate: 195 $/secWorldwide. No official global total. Scaled from US FTC/FBI figures; FTC Jan-Sep 2025: $1.16B reported (+22% vs same period 2024), annualized ~$1.54B US; FBI IC3 2025 covers all cybercrime with $20.8B total. US assumed to represent ~20-25% of global losses; converted to average dollars per second.
Rate: 1.018.519 /secWorldwide. ~1.1 billion people use ad blockers, blocking ~80 ads per day on average = 88.0 billion ads ads blocked per day. Converted to ads per second.
Rate: 3.472.222 /minWorldwide. ChatGPT alone ~2.5B requests/day (OpenAI July 2025); with Claude, Gemini, Copilot etc. conservatively ~4-5B/day. Converted to requests per minute.
Rate: 318,5 /secWorldwide. Modelled estimate: ~5 billion AI requests/day × 9.2% hallucination rate (Vectara lab benchmark for summarisation tasks) = 27.5 million false answers estimated false answers/day. Note: 9.2% is a benchmark, not a measured production rate; real-world rates vary widely by task and model.
Rate: 2.130 $/secWhen AI outputs are wrong or fabricated, the damage is real. Companies using AI for research, legal work, finance, or customer service pay for bad decisions, rework, lawsuits, and reputational hits. F
Rate: 380 /hrWorldwide. Resemble AI 2025: ~8 million deepfakes circulating online (16x vs 2023); annual new creation extrapolated to ~3.3 million/year ÷ 365 ÷ 24 ≈ 380/hr. 98% of deepfakes are non-consensual intimate imagery; 99% of victims are women (Security Hero 2024).
Rate: 3.000.000.000 /dayWorldwide. Estimated 3.0 billion posts false or misleading social media posts per day, based on ~7.5 billion daily posts (DataReportal 2024) and a 40% high-end estimate for misleading content. Conservative estimates (10–15%) would imply ~750M–1.1B/day. MIT/Science 2018: false news spreads 6× faster than true news.
Rate: 50.486 /dayWorldwide. Estimated ~50,486 songs songs per day that match the 'nobody will ever hear' definition: Luminate daily new ISRCs (~106,000/day in 2025) × share of catalog receiving fewer than 10 streams in an entire year (120.5M of 253M tracks ≈ 47.6%, Luminate 2025 Year-End Report).
Rate: ~47 sec average attention span on computer tasks (Ohio State survey, May 2025, n=1,000 US adults)Worldwide. Average adult attention span in digital work environments ~47 seconds seconds (Gloria Mark, UC Irvine, 'Attention Span', 2023); down from ~150 sec (2004) and ~75 sec (2012).
Rate: 0,08 /secPer person. Estimates of 4,000-10,000 ad exposures per person per day. Midpoint ~7,000 per day; counter shows how many ads the average person has been exposed to so far today.
Rate: 159 minutes/dayPer person. DataReportal/GWI 2026: ~2 hours 39 minutes (159 min) per day per social media user (18 h 36 min per week, 54 countries, ages 16–64; includes video-centric feeds like YouTube and TikTok).
Rate: 82.200 /dayWorldwide. U.S. data (PNAS 2019 Rosenfeld et al., The Knot 2025): ~2.4M marriages/year, 60% met online. Scaled to global ~50M marriages/year; applying similar share gives ~30M couples per year who met online = 82,200 couples per day.
Rate: 1,2 /minNew roles linked to automation and technology, worldwide. Modelled estimate derived from BLS 2024 tech hiring data, scaled globally. No official source tracks automation job creation globally; the rate (~1.2/min) is a conservative lower-bound model. WEF's newer Future of Jobs 2025 projects 170M new jobs by 2030, which is a forecast, not the live counter rate.
Rate: 20 /minDisplacement linked to automation and technology, worldwide. Modelled estimate: BLS displaced worker data (6.3M US workers 2021–2023) scaled globally with a conservative automation share. Rate (~20/min) is a modelled estimate; no global displacement count exists. WEF Future of Jobs 2025 projects 92M displaced by 2030 (~18M/year), which is a forecast, not the live counter rate.
Rate: 228 /hrWorldwide. WEF Future of Jobs Report 2025: ~2 million jobs displaced by automation per year (2024 baseline). MIT Acemoglu & Restrepo confirm causal link.
Rate: 11 /hrWorldwide. Based on WHO global road traffic deaths (approx. 1.19 million per year) and conservative shares from official distracted-driving statistics. Actual numbers vary by country and year.
Rate: 360 /hrPeople injured in crashes linked to digital distractions (e.g. texting, scrolling), worldwide. Rate scaled from NHTSA U.S. data (315,167 injured in 2024) to global road traffic; many countries underreport.
Rate: 6 /dayU.S. only. NHTSA Standing General Order: incidents where an SAE Level 2+ automated/driver-assistance system was active. 2024: 1,384 incidents (3.8/day). 2025 partial (Jan–Nov 17): 1,793 incidents (~5.6/day); the live counter uses this most recent rate. Note: 'involving' an active system ≠ caused by the system.
Rate: 40.509 /secWorldwide. Based on APWG 2025 campaign trajectory: ~3.5 billion phishing emails per day. 3,500,000,000 ÷ 86,400 ≈ 40,509/sec. This measures mass-blast and spear-phishing email volume; APWG separately tracks unique phishing sites and campaigns (a different metric).
Rate: 2.220 kg/secWorldwide. E-waste tracked in millions of tonnes per year globally. 62 million tonnes in 2022. Only 22.3% documented as recycled. Converted to kg per second.
+0,02 kg per second. That's ~1 kg per minute.≈ 560 tonnes of gold are lost in e-waste in 2025 (interpolated on UN GEM 2024 trajectory), worth about $55 billion at current gold prices.
Rate: 73.920.000.000 iPhone charges/minWorldwide. Data centres used ~485 TWh in 2025 (IEA, Apr 2026), about 1.5-1.7% of global electricity. Shown as equivalent household-days of consumption (1 household-day = 30 kWh).
Rate: 1.808.219.178 L/dayWorldwide. Global data center water consumption estimated at ~660 billion liters per year (2025 estimate, Nature Machine Intelligence / UC Riverside). That is ~1.81 billion liters per day ÷ 86,400 seconds ≈ 20,900 liters/second. Individual hyperscale facilities use 11–19 million liters/day; most of the world's ~8,000 data centers are far smaller.
Rate: 1.525.185 /secWorldwide. Billions of emails are sent every day; a large share are never read (~35% unread). Converted to unread emails per second.
Rate: 5.006.944 /minWorldwide. The "per minute" rate is derived from device count and typical charging frequency (~once per day per device). Converted to charge sessions per minute.
Rate: 805.479 /dayWorldwide. DataReportal 2025: net ~805,479 new internet users per day (year-on-year user growth). This is net annual growth, not absolute lifetime first-timers – it includes new users but may also reflect some who lapsed and returned.
Rate: 395 TByte/secWorldwide. AppLogic Networks (formerly Sandvine) 2025 Global Internet Phenomena Report: ~34 exabytes of data flow across the internet every day, roughly 395 TB/sec. This is a model-based carrier-network estimate; directly measured Internet Exchange traffic (~79 EB/year at major IXPs) is a verifiable subset at ~2.5 TB/sec.
Rate: 150 /minWorldwide. Hard drives have a measurable annual failure rate (AFR). Extrapolated from global installed drive base to failures per minute.
Rate: 380 /minWorldwide. ~200 million voice-enabled smart home devices (smart speakers, smart displays, and voice-enabled hubs) shipped per year (Statista/TechInsights 2024). Converted to ~380 devices per minute. Pure smart speaker shipments are ~87 million/year; the counter uses the broader voice-device category.
Rate: 2,5 /hrWorldwide. Reports submitted to reporting centers, not verified events. Rate scaled from NUFORC/MUFON U.S. reporting volumes to a global estimate. Converted to an average per hour.
Rate: 276.551 /dayWorldwide. DNIB Q4 2025: ~386.9M domain registrations; ~26.1% are not renewed annually. Estimated 276,551 domains domain expirations per day (all TLDs, including parked and never-active domains).
All numbers here come from public annual reports (WHO, FBI, DataReportal, and others) and are converted into continuously updated figures you can check against the original source. We document sources and how we calculate everything on the methodology page.
AnythingCounter converts published research into continuously updated figures. A yearly phishing total becomes a running count of today's emails. An annual e-waste report becomes a growing weight in kilograms. The source numbers always come from official reports and peer-reviewed research, not guesswork.
The idea behind it: big annual statistics are hard to grasp. Seeing them accumulate by the second makes the scale easier to understand. More about the project.
What you can find here
The counters cover AI hallucinations, deepfakes, phishing emails, e-waste, jobs displaced by automation, surveillance cameras, and more across 10 categories. Each statistic has its own page with historical data, sources, and context. Browse all statistics.
How the numbers are calculated
We take annual or periodic figures from sources like the WHO, FBI, UN, and DataReportal and convert them into a per-second rate. Your browser then multiplies that rate by the time that has passed since midnight or since January 1st. The underlying rates are fixed until a newer study is published, at which point they get updated manually. For the full calculation logic and source list, see the methodology page.
Who uses it and what for
Mostly people who want to put a number in context. A journalist writing about cybercrime, a teacher explaining AI risks, a designer looking for a striking data point. Watching a number tick up by the second communicates scale in a way that a table of annual figures often does not. All statistics link back to the original source, so you can verify and cite them directly.
AnythingCounter FAQ
What is AnythingCounter?
AnythingCounter converts published statistics on AI, deepfakes, phishing, e-waste, and automation into per-second and per-day numbers. Instead of reading that phishing costs businesses “billions a year,” you can see roughly how many emails go out while you're reading this page.
How are the statistics calculated?
We use published yearly or periodic stats from international bodies, research institutes, and cybersecurity reports. Those figures get converted into per-second or per-day rates and shown as live counters. Full sources and how we calculate everything: our methodology page.
Are the numbers real or estimated?
They're modelled from real, cited data, not guesses. A published yearly or industry total gets converted into a steady rate, and every figure links back to the original source so you can check it.
How often do the counters update?
The totals recalculate constantly based on the underlying rate. The rates themselves come from the newest research we could find, and we revise them whenever better data is published.
Data sources: Amra & Elma 2024, Anthropic, APWG, Axios, Backblaze Drive Stats, BLS American Time Use, BLS Displaced Workers, BLS Employment, Cisco VNI, Cloudflare, Colorlib, Cornell, CyberPeace 2023, Cyberbullying Research Center, DataReportal, DemandSage, DE-CIX, Digital Silk, Domain Name Industry Brief (DNIB), EIA, Eksposure, ENERGY STAR, Everypixel, FDM Group 2024, Fakespot, FBI IC3 2024, Feedzai, Forrester Research 2024, Freedom House 2024, FTC, G2, GASA Global State of Scams 2024, Google AI, Hostinger, IFR World Robotics 2025, INTERPOL, ITU E-waste, Legiscope, Luminate, Microsoft Research, Microsoft 2025, Music Business Worldwide 2024, MUFON, NASA Astrobiology, Netcraft, NewsCatcher, NHTSA, NUFORC, OpenAI, Our World in Data, PageFair, Pangram Labs, Photutorial, PNAS 2024, Science (MIT), Security Hero 2024, seosandwich.com 2024, Social Rails, Spotify 2024, Statista, Statusbrew, The Knot 2025, The Times, TechCrunch 2025, UN Global E-waste Monitor, US Geological Survey, Vectara Hallucination Leaderboard 2025, VeriSign, Website Rating 2024, WEF Future of Jobs 2025, WHO, YouTube
