Google published a guide titled "Allow Google-Agent traffic" on November 17, 2026, telling website operators and their CDN, firewall and bot-detection vendors how to let its AI agent through, PPC Land reported. The guide sets four rules: confirm the User-Agent header contains Google-Agent, confirm the IP address is in Google's user-triggered-agents.json file, verify a Web Bot Auth signature if one is provided, and skip bot challenges for requests that pass.
The guide lists six checks for when legitimate requests are still blocked, including rule priority, interactive challenges and per-IP rate limits. Google says concurrent requests from multiple users share the same egress pool, and that broad data center blocks, managed bot rules or category-level AI agent block policies can override an allow rule. Its list of in-journey actions includes on-site search, filtering, add-to-cart and booking. Google's fetcher page says such user-triggered fetchers generally ignore robots.txt rules, and only a subset of Google-Agent requests is signed.
PPC Land noted the guide gives no volume figures, names no vendors or Google products that use the agent, and does not explain how to block it. The outlet cited Blend's audit, which found bot protection blocked its AI shopping agent at 29 of 200 UK and US retailers. Separately, PPC Land reported that Google made Gemini's task mode, which can compare products and check out, free in six countries the same day.






