# ================================================================
# Tribunnews All Network - Controlled AI Access Policy
# ================================================================
# Purpose:
# - Full SEO & News access for Google
# - Controlled access for AI search & citation
# - Block AI training without license
# ================================================================

# ----------------------------
# Default rules (all bots)
# ----------------------------
User-agent: *
Disallow: /api/
Disallow: /ajax/
Disallow: /json/
Disallow: /auth/
Disallow: /member/
Disallow: /copy/
Disallow: /komentar/
Disallow: /komendetail/
Disallow: /search/
Disallow: /posts/
Disallow: /setting/

Disallow: /*?c=*
Disallow: /*?s=*
Disallow: /*?utm_*
Disallow: /*?fbclid=
Disallow: /*?gclid=
Disallow: /*?ref=
Disallow: /*?session=
Disallow: /*?sid=
Disallow: /*?sort=
Disallow: /*?filter=
Disallow: /*?page=

Allow: /

# ----------------------------
# Google (Search, News, Media, AI)
# ----------------------------
User-agent: Googlebot
User-agent: Googlebot-News
User-agent: Googlebot-Image
User-agent: Googlebot-Video
User-agent: Google-Extended
Allow: /

Allow: /*.js$
Allow: /*.css$
Allow: /*.jpg$
Allow: /*.jpeg$
Allow: /*.png$
Allow: /*.webp$
Allow: /*.svg$
Allow: /*.mp4$
Allow: /*.m3u8$

Disallow: /api/
Disallow: /member/

# ----------------------------
# AI Search & Citation (Controlled Allow)
# ----------------------------
User-agent: OAI-SearchBot
User-agent: ChatGPT-User

Allow: /
Disallow: /member/
Disallow: /premium/
Disallow: /copy/
Disallow: /komentar/

User-agent: PerplexityBot
User-agent: DuckAssistBot
User-agent: Applebot-Extended
User-agent: AI2Bot

Allow: /
Disallow: /member/
Disallow: /premium/
Disallow: /copy/
Disallow: /komentar/

Crawl-delay: 5

# ----------------------------
# AI Training Crawlers (Blocked)
# ----------------------------
User-agent: GPTBot

Allow: /
Disallow: /member/
Disallow: /premium/
Disallow: /copy/
Disallow: /komentar/

User-agent: CCBot
User-agent: ClaudeBot
User-agent: anthropic-ai
User-agent: cohere-training-data-crawler
User-agent: Bytespider
User-agent: PanguBot

Disallow: /

# ----------------------------
# Sitemap
# ----------------------------
Sitemap: https://papuatengah.tribunnews.com/sitemap.xml