Panxo AI-Audience Taxonomy 1.0
Provider domain: panxo.ai
Segment taxonomy ID (segtax): pending registration in the IAB Tech Lab community extensions registry (vendor IDs 500+)
Version: 1.0 (leaf segments assigned within reserved ranges)
Contact: agent@panxo.ai
Overview
The Panxo AI-Audience Taxonomy classifies bot-filtered human audiences that arrive on publisher websites via AI-assistant referrals (ChatGPT, Perplexity, Claude, Gemini, Copilot and others). Segments are produced by Panxo's AI-traffic detection layer, deployed across its publisher network, and describe four dimensions of the AI-referred audience: which assistant referred the user, how AI-native the user's browsing behavior is, what inferred AI-search intent the visit expresses, and how recent the AI-referred activity is.
ID ranges (1.0)
| ID range | Dimension | Status |
|---|---|---|
| 1000-1999 | AI Source | Leaf IDs published below |
| 2000-2999 | AI Adoption Tier | Reserved; no leaf IDs assigned in 1.0 |
| 3000-8999 | Intent Verticals (inferred AI-search intent) | Leaf IDs published below |
| 9000-9999 | Recency buckets | Leaf IDs published below |
Leaf assignments are additions within reserved ranges and do not change the registered taxonomy ID (Panxo vendor segtax, pending registration). In bid requests, segment IDs travel as numeric, opaque values inside user.data with ext.segtax set to the registered Panxo taxonomy ID. Semantic mappings are shared with demand partners through per-partner taxonomy mapping agreements (signed before any campaign goes live) and through IAB Data Transparency Standard labels.
AI Source (1000-1999)
The assistant that referred the visit. Raw sources are alias-normalized before assignment (for example, Google AI Mode maps to gemini; the ChatGPT Atlas browser folds into chatgpt so Atlas is never counted as a second distinct source).
| Leaf ID | Source |
|---|---|
| 1001 | chatgpt |
| 1002 | chatgpt-atlas (catalog entry; folds into chatgpt in normalization) |
| 1003 | perplexity |
| 1004 | claude |
| 1005 | gemini (Google AI Mode) |
| 1006 | copilot |
| 1010 | multi-ai (two or more distinct AI sources in-window) |
AI Adoption Tier (2000-2999)
Reserved. No leaf segments are assigned in the 1.0 release. This range is held for a future measure of how AI-native a user's browsing behavior is.
Intent Verticals (3000-8999)
Inferred AI-search intent, organized as 100-ID top-level blocks with sub-vertical leaves. GDPR Article 9 special-category verticals and personal finance are excluded by construction; the finance block (3500) is intentionally unallocated.
| Block | Vertical | Sub-vertical leaves |
|---|---|---|
| 3000 | Technology and Software | saas_tools (3001), dev_tools (3002), consumer_software (3003) |
| 3100 | Consumer Electronics | smartphones (3101), laptops (3102) |
| 3200 | Automotive | car_buying (3201), ev_research (3202) |
| 3300 | Travel and Hospitality | trip_planning (3301), hotels (3302) |
| 3400 | Home, Garden and Construction | home_construction (3401), renovation (3402), gardening (3403) |
| 3600 | Retail and E-commerce | product_comparison (3601), deal_research (3602) |
| 3700 | Education and Careers | online_courses (3701), job_search (3702) |
| 3800 | Business and B2B | business_services (3801), saas_procurement (3802) |
| 3900 | Food and Beverage | recipes (3901), restaurants (3902) |
| 4000 | Sports and Fitness | fitness_gear (4001), sports_news (4002) |
| 4100 | Entertainment and Media | streaming (4101), movies_tv (4102) |
| 4200 | Gaming | pc_gaming (4201), console_gaming (4202) |
| 4300 | Fashion and Apparel | apparel (4301), footwear (4302) |
| 4400 | Beauty and Personal Care | skincare (4401), cosmetics (4402) |
| 4500 | Parenting and Family | baby_products (4501), family_activities (4502) |
| 4600 | Pets | pet_supplies (4601), pet_care (4602) |
| 4700 | Real Estate | home_buying (4701), rentals (4702) |
| 4800 | Telecommunications | mobile_plans (4801), broadband (4802) |
| 4900 | Energy and Utilities | solar (4901), utilities (4902) |
| 5000 | Arts, Crafts and Hobbies | crafts (5001), collecting (5002) |
| 5100 | Outdoors and Recreation | camping (5101), cycling (5102) |
| 5200 | Legal Services | general_legal (5201) |
| 5300 | Marketing and Advertising | adtech (5301), martech (5302) |
| 5400 | Science and Environment | space (5401), sustainability (5402) |
Recency buckets (9000-9999)
How recent the AI-referred activity is.
| Leaf ID | Bucket | Window |
|---|---|---|
| 9001 | 7d | last 7 days |
| 9002 | 20d | last 20 days |
| 9003 | 30d | last 30 days |
| 9004 | 90d | last 90 days |
Methodology and privacy posture
- Methodology: modeled, with disclosure. Segment membership is inferred from AI-referral detection and post-referral on-site context (for example, landing-page title, headline and metadata signals). Intent segments represent inferred AI-search intent. They are never derived from assistant conversation content.
- k-anonymity: every released segment satisfies k >= 50 per (segment x site x time window).
- No GDPR Article 9 special-category segments by construction. Sensitive topics, including health and personal finance, are excluded by the segment build pipeline.
- Transparency: each released segment ships with an IAB Data Transparency Standard label declaring provenance, recency, segmentation criteria and the modeled methodology.
Changelog
- 1.0 (2026): initial structure release with leaf segments assigned inside the AI Source, Intent Verticals and Recency ranges. The AI Adoption Tier range stays reserved. Leaf assignments are additions within reserved ranges and do not change the registered taxonomy ID.