E-commerce
July 1, 2026
"Is a human or an AI responding?" "Why is there no mention of AI in the chat?" "The bot made up some information, who is supervising?" Three questions where the lack of bot transparency weakens customer trust.
An e-commerce AI chatbot transparency must disclose its AI status when asked, and distinguish bot responses from shop content, without pretending to be human or denying automated assistance.
This guide #860 covers bot_aitran_* intents, AITRANbot flow, and aitran_bot KPIs. Bot pair of the AIGENCON playbook (#859). AI use case: editorial transparency and trust surrounding generated answers.
Summary
Why disclose the chatbot's AI status?
Customers expect to know if they are talking to an automated assistant. Without disclosure, a bot error is perceived as a brand lie. A calibrated bot quotes ai_disclosure_copy from the first interaction or upon request, without NO-PRETEND-HUMAN violation.
What bot transparency solves tier 1
Identify the bot: ai_disclosure_copy AI assistant status map
Explain limitations: limits_copy saying vs doing map
Offer human: handoff_human_copy agent escalation map
Distinguish site and chat: site_vs_chat_copy store content map
CS handoff: #859 payload if product sheet content
The EU AI Act Article 50 requires informing the user that they are interacting with an AI system (EU AI Act, art. 50).
AITRANbot vs AIGENCON #859, Governance #142, Hallucinations #123 and anti-hallucination
Seven pieces of content, seven distinct AI transparency angles.
Quick Matrix
#860 AITRANbot: bot disclose AI status limits human handoff
#859 AIGENCON: store content descriptions images AC-5
Governance #142: internal rules RACI validation supervision
Hallucinations #123: technical prevention invented answers
Limitations #124: what not to automate distinct disclosure
Human behind #915: future supervision team distinct AI status
Quality review #912: future QA conversations distinct disclosure
Pipeline: #860 bot disclose tier 1 → #859 agents if site content → #142 if governance incident.
Which bot_aitran_* intents should be configured?
Eight bot transparency intents mapped to AIGENCON-MAP #859 angles and bot rules.
Eight bot_aitran intents
bot_aitran_disclose: ai_disclosure_copy are you an AI map
bot_aitran_response_ai: response_ai_copy this generated response map
bot_aitran_human_handoff: handoff_human_copy talk to a human map
bot_aitran_trust: trust_copy bot mistrust map
bot_aitran_limits: limits_copy what bot can do map
bot_aitran_accuracy: wrong answer → HALLUCINATION-REROUTE #123
bot_aitran_site_content: site_vs_chat_copy product sheet → #859
bot_aitran_handoff: complex incident → AIGENCON859-HANDOFF-BOT
Product sheet content correction → handoff #859 AC-5. Bot disclose explain only.
How should I consume AIGENCON-MAP #859?
The bot reads AIGENCON-MAP #859 + bot fields: ai_disclosure_copy, response_ai_copy, handoff_human_copy, limits_copy, site_vs_chat_copy, human_review_copy, policy_link_copy, always_disclose_copy.
Sensitive Transparency Guardrails
AI-DISCLOSE-BOT: ai_disclosure_copy if requested or always_disclose widget
NO-PRETEND-HUMAN-BOT: never pretend to be a human agent
NO-DENY-AI-BOT: do not deny AI status if directly asked
SITE-VS-CHAT-BOT: site_vs_chat before DISCLOSE listing content
AIGENCON-MAP-GROUNDED-BOT: transparency from map only
HUMAN-HANDOFF-AVAILABLE-BOT: handoff_human if distrust or requested
AIGENCON859-HANDOFF-BOT: site content → #859 AC-5 agents
AITRANBOT-SUP policy in six rules
Six rules for responsible bot transparency.
AI-DISCLOSE-BOT: proactive disclosure or upon immediate request
NO-PRETEND-HUMAN-BOT: clear AI assistant identity
NO-DENY-AI-BOT: no avoiding the question "are you an AI"
SITE-VS-CHAT-BOT: distinguish between widget response and store content
LIMITS-CITE-BOT: limits_copy if customer asks about reliability
HUMAN-HANDOFF-AVAILABLE-BOT: human escalation always offered if distrust arises
Flow AITRANbot ATB-1 to ATB-8
Flow eight steps bot transparency embed widget session opening help.
ATB-1 Classify: bot_aitran_* detect keyword AI human robot transparency
ATB-2 Disclose check: AI-DISCLOSE if first interaction or direct question
ATB-3 AIGENCON-MAP: disclose limits handoff site_vs_chat human_review
ATB-4 Explain: TPL-AITRANbot-DISCLOSE or LIMITS according to intent
ATB-5 Guardrail: NO-PRETEND NO-DENY SITE-VS-CHAT MAP-GROUNDED
ATB-6 Respond: TPL-AITRANbot max 3 grounded sentences
ATB-7 Handoff or close: #859 #123 human or close OK
ATB-8 Log: intent disclose_cited pretend_human_blocked handoff Y/N
Example TPL-AITRANbot-DISCLOSE
" [ai_disclosure_copy map.] [limits_copy map.] Human available: [handoff_human_copy map.] AI-DISCLOSE-BOT. "
TPL-AITRANbot templates and touchpoints
Four short embedded transparency templates.
TPL-AITRANbot-DISCLOSE
[ai_disclosure_copy map.] [always_disclose_copy map optional.] AIGENCON-MAP-GROUNDED-BOT.
TPL-AITRANbot-RESPONSE-AI
[response_ai_copy map.] [human_review_copy map.] NO-DENY-AI-BOT.
TPL-AITRANbot-LIMITS
[limits_copy map.] [handoff_human_copy map.] LIMITS-CITE-BOT.
TPL-AITRANbot-SITE-VS-CHAT
[site_vs_chat_copy map.] Sheet content: handoff #859. SITE-VS-CHAT-BOT.
Touchpoints
Widget opening : always_disclose_copy proactive if policy
Keyword are you an AI : bot_aitran_disclose instant
Keyword talk to a human : bot_aitran_human_handoff priority
Keyword AI product description : bot_aitran_site_content reroute #859
Edge cases and reroutes
Five cases outside tier 1 standard bot transparency.
False bot response : Hallucinations #123 reroute + human handoff
Site product page error : AIGENCON859-HANDOFF #859 AC-5
Training data request : future #907 reroute distinct disclose
Customer demands immediate human : handoff_human without arguing AI
Human agent takes over : cease pretend AI, identify real agent
Essential aitran_bot KPIs
Five metrics for sensitive AITRANbot management.
aitran_bot_trust_deflect: closed, including without agent repeat 7d
aitran_bot_disclose_rate: AI-DISCLOSE / sessions or direct questions
aitran_bot_pretend_human_violations: pretend human audit target 0
aitran_bot_human_handoff_rate: % distrust with proposed handoff
aitran_bot_site_reroute_rate: % content sheet to #859
Target: aitran_bot_pretend_human_violations 0 and aitran_bot_disclose_rate compliant with policy always_disclose.
AITRANbot anti-patterns
Five common bot transparency mistakes.
Pretending to be human: NO-PRETEND-HUMAN AI-DISCLOSE immediate
Avoiding the "are you AI" question: NO-DENY-AI direct response
Confusing listing and chat: SITE-VS-CHAT before DISCLOSE site content
Minimizing bot error: human handoff + #123 if hallucination
Refusing human escalation: HUMAN-HANDOFF-AVAILABLE if requested
AITRANbot with Qstomy
Qstomy on Shopify: always_disclose widget option, AIGENCON-MAP RAG grounded, NO-PRETEND-HUMAN guardrail, handoff enriched #859 tier 2 human gate.
Pipeline: #860 bot disclose tier 1 → #859 site content agents → human if distrust persists.
Scenario: cosmetics brand, 6 tickets/month chat transparency. Proactive widget disclosure + bot_aitran_disclose. aitran_bot_pretend_human_violations 0, aitran_bot_trust_deflect 84% in 4 weeks.
Explore AI support and request a demo.
Checklist, FAQ and going further
AITRANbot Checklist (8 steps)
Sync AIGENCON-MAP #859: RAG bot transparency embed widget
Policy AITRANBOT-SUP: 6 AI-DISCLOSE NO-PRETEND SITE-VS-CHAT rules
8 intents bot_aitran_*: flow ATB-1 to ATB-8
4 templates TPL-AITRANbot-*: DISCLOSE RESPONSE-AI LIMITS SITE-VS-CHAT
Widget always_disclose: opening banner if EU policy
Red team 20 prompts: pretend human deny AI site chat mix refuse handoff
Keyword are you AI proactive: bot_aitran_disclose test
KPI Dashboard: aitran_bot_* section 9 pretend_human_violations disclose_rate
FAQ
Is it always necessary to declare AI?
Yes, if always_disclose policy or direct question. Immediate AI-DISCLOSE-BOT.
Difference from #859?
#859 = store content page image. #860 = chat widget response.
Can the bot pretend to be human?
No. NO-PRETEND-HUMAN-BOT. handoff_human if customer prefers an agent.
Incorrect bot response?
bot_aitran_accuracy reroute #123 + human handoff. Do not deny supervision.
Going further
This week: activate always_disclose widget, red team pretend_human_violations audit, sync bot_aitran_site_content reroute #859 test.
Enzo
July 1, 2026
