# Privasee robots policy
#
# Two intents, two policies:
#   1. Training bots (ingest content to train LLMs)      → BLOCKED (EU DSM Art. 4 reservation)
#   2. Inference/search bots (cite our pages in answers) → ALLOWED (we want AI search visibility)
#
# Rationale : we oppose ingestion of our content for model training, but we
# want to be discovered and cited by AI-powered search (ChatGPT, Claude,
# Perplexity, Gemini) when users ask about DLP / PII detection / GDPR.

# ── Traditional search crawlers ───────────────────────
User-agent: Googlebot
Allow: /

User-agent: Bingbot
Allow: /

User-agent: *
Allow: /

# ── AI inference / search crawlers (ALLOWED) ──────────
# These crawlers fetch pages in real time to cite them in AI answers.
# They do NOT ingest content for training.

# Perplexity — real-time retrieval for cited answers
User-agent: PerplexityBot
Allow: /

# OpenAI — ChatGPT live browsing (triggered by user queries)
User-agent: ChatGPT-User
Allow: /

# OpenAI — ChatGPT search index (public search surface)
User-agent: OAI-SearchBot
Allow: /

# Anthropic — Claude real-time browsing (user-initiated)
User-agent: Claude-User
Allow: /

# Anthropic — Claude public search index
User-agent: ClaudeBot-User
Allow: /

# Google — Gemini / AI Overviews grounding. ALLOWED on purpose: unlike the pure
# training crawlers below, Google-Extended also gates whether our pages can be
# cited in Gemini and Google AI Overviews. Blocking it would suppress AI-search
# visibility (our goal) without meaningfully protecting public marketing content.
User-agent: Google-Extended
Allow: /

# ── AI training / broad scraping (BLOCKED) ────────────
# These crawlers ingest content to train or fine-tune language models.
# Blocking aligns with EU DSM Directive 2019/790 Article 4 reservation.

# OpenAI — training crawler
User-agent: GPTBot
Disallow: /

# Anthropic — training crawler
User-agent: ClaudeBot
Disallow: /

# Anthropic — legacy training UA
User-agent: anthropic-ai
Disallow: /

# Apple — Apple Intelligence training opt-out
User-agent: Applebot-Extended
Disallow: /

# Common Crawl — dataset used by many foundation models
User-agent: CCBot
Disallow: /

# ByteDance — used for training
User-agent: Bytespider
Disallow: /

# Amazon — ambiguous intent, blocked by default
User-agent: Amazonbot
Disallow: /

# Meta — training + link expansion
User-agent: meta-externalagent
Disallow: /

User-agent: FacebookBot
Disallow: /

# Cohere — training
User-agent: cohere-ai
Disallow: /

# Allen Institute — training
User-agent: AI2Bot
Disallow: /

# Diffbot — commercial scraping
User-agent: Diffbot
Disallow: /

# Generic scrapers
User-agent: Scrapy
Disallow: /

User-agent: MJ12bot
Disallow: /

User-agent: AhrefsBot
Disallow: /

User-agent: SemrushBot
Disallow: /

User-agent: DotBot
Disallow: /

# ── Legal reservation (EU DSM Directive 2019/790 Art. 4) ─────
# No authorization granted for AI model training ingestion of this site's
# content. Any training-intent crawler accessing this site does so in
# violation of the rights reservation expressed above.

Sitemap: https://privasee.fr/sitemap.xml