Every AI bot that visits your store — and whether you should let it in.
ARC runs five browser shopper personas (ARC browser shoppers A to E) on ARC-chosen models. Personal agents shop by fetching your pages. These are the named crawlers they rely on. Some you want. Some you may not.
Want to know which of these can reach your store right now?
Run a free scan →AI assistants answering shoppers
These fetch your page the moment a shopper asks a personal agent about you. Blocking them makes you invisible — the costliest block.
Fetches your page when a ChatGPT user asks about it. Block this and ChatGPT can't read you for a shopper.
SHOULD YOU ALLOW IT? →Fetches your page when a Claude user asks about it. Block this and Claude can't answer questions about your products.
SHOULD YOU ALLOW IT? →Perplexity's live fetch for answering user questions and 'Buy with Pro'.
SHOULD YOU ALLOW IT? →User-triggered fetch when a Perplexity user opens your page.
SHOULD YOU ALLOW IT? →Google's assistant reading your page to answer a live query (governed by Google-Extended).
SHOULD YOU ALLOW IT? →Microsoft Copilot / Bing Chat reading your page to answer and shop.
SHOULD YOU ALLOW IT? →Amazon's assistant (Rufus / Alexa) fetching product info.
SHOULD YOU ALLOW IT? →Meta AI (WhatsApp / Instagram) fetching your page on a user's behalf.
SHOULD YOU ALLOW IT? →AI search indexing you
These index you so AI search can surface you. Blocking them means you won't appear in AI results.
Indexes your site so ChatGPT search can recommend it. Block this and you vanish from ChatGPT's results.
SHOULD YOU ALLOW IT? →Indexes your site for Claude's search results.
SHOULD YOU ALLOW IT? →Google's core crawler — also feeds AI Overviews and AI Mode.
SHOULD YOU ALLOW IT? →Bing's crawler — feeds Copilot's shopping answers.
SHOULD YOU ALLOW IT? →DuckDuckGo's assistant index.
SHOULD YOU ALLOW IT? →Feeds Siri / Apple Intelligence answers.
SHOULD YOU ALLOW IT? →Model training crawlers
These crawl to train models. Blocking them is a legitimate content choice and does NOT stop AI assistants from shopping you live.
Crawls to train OpenAI models. Blocking is a valid content decision — it does NOT affect whether ChatGPT can shop you live.
SHOULD YOU ALLOW IT? →Crawls to train Anthropic models.
SHOULD YOU ALLOW IT? →Opt-out control for training Gemini on your content.
SHOULD YOU ALLOW IT? →Crawls to train Meta's Llama models.
SHOULD YOU ALLOW IT? →Amazon's training crawl.
SHOULD YOU ALLOW IT? →ByteDance (TikTok) training crawler — aggressive; often blocked.
SHOULD YOU ALLOW IT? →Common Crawl — the open dataset most LLMs train on.
SHOULD YOU ALLOW IT? →