• 2 min read
Meta bots hit 9 billion requests with no publisher traffic
DataDome says Meta AI bots made 9 billion requests in Q2 2026, up sharply from Q1, while sending publishers virtually no traffic.
Image: TechRadar
A DataDome analysis says Meta AI bots generated 9 billion requests in Q2 2026, pushing up publishers' bandwidth, logging, resource, and CDN costs while returning no meaningful traffic in exchange. The bot management and agent control company says broader agentic traffic rose 45% in the quarter, with Meta AI bots up more than 163% from Q1.
For website operators, that changes the usual crawler tradeoff. Traditional search engines consumed resources too, but they also sent visitors back. According to DataDome, AI bots increasingly do not.
Meta-WebIndexer growth and ChatGPT referrals
A key signal in the report is the rise of Meta-WebIndexer, which DataDome says suggests Meta may be building a web index similar to the ones search engines have maintained for decades. That would help explain the heavy crawl volume seen in Q2 2026.
Recommended reading
Agility opens Fremont robot hub near Tesla’s Optimus push
By contrast, ChatGPT crawlers reportedly declined in traffic while delivering more value to publishers. TechRadar says ChatGPT accounts for 88% of AI referrals, making it comparatively efficient even with fewer crawls.
DataDome also flagged the emergence of Model Context Protocol (MCP) traffic, which connects AI agents to external tools and does not look like standard crawler behavior.
“Q2 showed us that the ground is shifting faster than most organizations realize. Meta now dominates AI traffic on our network, MCP traffic has emerged as a real signal, and ChatGPT is driving more referral value with fewer crawls.”
How sites may respond
DataDome frames this partly as a security and infrastructure problem. Different agents behave differently: some act more like users, others scrape content, and unknown agents may present greater risk.
That could push organizations to set more granular policies, such as:
- allowing full crawl access to Google
- permitting ChatGPT retrieval
- rate-limiting Meta AI
- requiring extra verification for unknown agents or blocking them outright
Segura said companies that distinguish between these agent types are gaining an advantage as agent trust tools spread.
Enterprise Editor
Marcus follows the money. He covers enterprise software, cloud architecture, and the tectonic shifts in Big Tech strategy. He translates dense earnings calls and complex M&A activity into actionable insights about where the industry is actually heading. If a tech giant makes a silent pivot, Marcus is usually the first to notice.
via TechRadar
