More Than Half Of Web Traffic Is Bots. Ads Can't Survive It

More than half of all web traffic is now bots — automated scripts and AI agents that buy, execute, and leave without ever seeing a single ad. The 30-year-old advertising revenue model that funded the open web is cracking, and payment giants Visa, Mastercard, and infrastructure provider Cloudflare are already positioning themselves as toll collectors for this new machine-to-machine economy.

The revelation comes from a Forbes analysis of traffic data and industry reports. Bots — including search engine crawlers, scrapers, and increasingly sophisticated AI agents from companies like OpenAI — now account for 50% to 60% of all internet requests. Human browsing, the foundation of digital advertising, is shrinking to a minority slice. The ad industry, which generated over $600 billion globally in 2024, relies on human eyeballs to justify its CPMs. When bots make up the majority of visits, impressions lose value, click-through rates become meaningless, and fraud balloons.

The problem isn't new — ad fraud has plagued the web for years. But the scale is unprecedented. In the past, most bots were simple scrapers or click farms. Today's AI agents, like ChatGPT's browsing mode or Perplexity's citation engine, behave like real users: they load pages, parse content, and interact. They just never load ads. Ad blockers and cookie opt-outs already suppressed revenue. AI agents are the final blow.

Visa and Mastercard have separately filed patents for systems that verify whether a request comes from a human or a bot before processing a transaction. The idea: charge a micro-fee for each bot visit to a publisher's site, effectively creating a toll for automated access. Cloudflare, which routes about 20% of global web traffic, already offers bot management tools and is testing a “request verification” layer that could become a default gatekeeper. If adopted broadly, every bot call could carry a tiny payment — fractions of a cent — that flows back to publishers.

Industry analysts see a fundamental shift. "The ad model isn't broken — it's being replaced," says digital economy researcher Dr. Elena Marchetti. "When machines are the majority of traffic, you need a machine-scale revenue model, not a human-scale one." The implications are huge: publishers could stop relying on ad networks entirely, instead charging API-style fees for access to their content by bots. That would upend Google's search business, which depends on crawling the open web for free, and reshape how AI companies train their models.

What happens next depends on how quickly the toll infrastructure rolls out. Visa and Mastercard's systems are in early trials. Cloudflare has hinted at a 2026 launch for its verification layer. Meanwhile, major publishers like The New York Times and Axel Springer have already cut deals with AI companies for training data. If tolls become standard, those deal values could skyrocket — or collapse if bots simply stop visiting sites that charge. The next two years will determine whether the web remains an advertising-driven public square or becomes a toll-road of pay-per-request transactions. One thing is clear: the era of free, ad-supported browsing for machines is ending.