Anti-bot systems increasingly flag LLM-driven browsers as bots, creating a new blocking frontier
ScrapingBee examines how anti-bot vendors like Cloudflare, DataDome, Akamai, and HUMAN are classifying and blocking browsers operated by language models, such as Perplexity's Comet, ChatGPT browsing, Claude computer use, and Playwright MCP agents.
Briefing
- ScrapingBee: ScrapingBee examines how anti-bot vendors like Cloudflare, DataDome, Akamai, and HUMAN are classifying and blocking browsers operated by language models, such as Perplexity's Comet, ChatGPT browsing, Claude computer use, and Playwright MCP agents.
Why it matters
As LLM-driven browsing becomes more common for research, automation, and AI agents, the arms race between bot operators and anti-bot systems is entering a new phase. This creates friction for legitimate use cases like personal AI assistants and automated data collection, forcing developers to adapt their browser fingerprints and request patterns to avoid being blocked. The trend also raises questions about how anti-bot vendors will distinguish between helpful AI agents and malicious scrapers.
Sources
- Primary source · ScrapingBee · 2026-10-02 — Bots Block Bots: Why Anti-Bot Systems Flag LLM Browsers
Watch next
Will anti-bot vendors introduce dedicated LLM-browser classifications that allow legitimate AI agents to pass while still blocking malicious scrapers?