At BoldDesk, data protection is a fundamental principle integrated into all aspects of our AI chatbot development. Below are responses to frequently asked questions from support teams regarding the collection, management, and protection of their data and their customers’ data. This document explains how BoldDesk AI manages, processes, and protects support content, customer conversation data, and AI‑generated interactions. Each section is designed as an atomic, standalone knowledge unit for maximum retrieval accuracy.

1. Data Processing Overview

BoldDesk follows a data-protection-first design approach. All AI-related processing adheres to strict privacy, security, and customer‑control principles. BoldDesk’s AI data processing is currently centralized in the United States, regardless of where your data is stored (for instance, EU regions).

2. How BoldDesk AI Processes Support Content

How your internal support material is used by BoldDesk AI.

• BoldDesk AI processes knowledge base articles when they are created or updated.
• This processing enables the AI to generate customer responses based on your documented support content.
• Specific ticket content is only processed when using features such as:
  • Translation
  • Review
  • Rephrasing
• Only the relevant input content you supply is accessed—nothing else.

3. How BoldDesk AI Handles Customer Conversation Data

How customer messages and AI conversations are processed.

• All customer prompts, messages, and conversation data sent to the AI are forwarded directly to Azure OpenAI for response generation.
• Customer interactions may include personal information voluntarily shared by the user.

4. Monitoring of Customer–AI Interactions

When and how interactions may be monitored.

• Certain AI features may use third‑party model providers to monitor prompts and AI responses.
• Monitoring is used only for:
  • Misuse detection
  • Harmful content detection
  • Safety and abuse prevention
• Monitoring can be:
  • Automated
  • Human‑assisted (human review)

5. Use of Customer Data for Model Training

BoldDesk explicitly ensures that your data:

• is NOT accessible to other customers
• is NOT shared with OpenAI
• is NOT used to train OpenAI models
• is NOT used to improve Azure OpenAI foundation models
• is NOT used for any third‑party product or service improvement without your consent

6. Data Retention by Azure OpenAI

Azure OpenAI does not retain or store any of your data.

7. Models Used by BoldDesk AI

BoldDesk AI uses Azure OpenAI language models, including:

• GPT‑5.2
• GPT‑5.4
• GPT‑5.4 Mini
• GPT‑5.4 Nano

These models provide fast, accurate, and conversational AI assistance for customer support interactions.

8. Controlling the Content Used by BoldDesk AI

You can configure which sources the AI uses for generating responses, such as:

• Your help center
• Custom knowledge base articles
• Publicly accessible URLs containing your support documentation

This ensures:

• Fine‑tuned responses
• Context‑accurate answers
• Ability to update or expand reference sources anytime

9. When BoldDesk AI Cannot Find an Answer

If BoldDesk AI cannot locate relevant information:

• It acknowledges the limitation instead of generating fabricated or inaccurate responses.
• It always references validated sources from your support content when producing answers.

10. Accuracy of AI Responses

• AI-generated responses are probabilistic and may occasionally contain inaccuracies.
• While Azure OpenAI GPT‑5 models improve response quality and reliability, incorrect or incomplete information can still occur, especially with:
  • Ambiguous or unclear queries
  • Missing, unavailable, or limited context and content sources

11. Does the AI Agent Respond Directly to Customers?

Direct AI responses are supported in:

• Chat through:
  • Livechat
  • WhatsApp (Meta Cloud API)
  • Facebook
  • Instagram
  • Twilio SMS
  • Telegram
  • WhatsApp (Gupshup)
  • Line
  • WhatsApp (Twilio). Explore How to Assign AI Agent 2.0 to Omnichannel in BoldDesk.
• Web Widget. Explore How to Enable AI 2.0 Agent in BoldDesk’s Web Widget.
• Tickets. Explore How to Auto-Assign AI 2.0 Agent for Ticket Responses.

To enable AI responses:

• Admins must create and configure the AI Agent in the AI Center. Explore How to Create and Manage AI 2.0 Agent in BoldDesk.

Frequently Asked Questions

1. Does BoldDesk use my data to train AI models?
   No. Your data is never used for training OpenAI or Azure OpenAI models.

2. Can I choose which knowledge sources the AI uses?
   Yes. You can select Knowledge base articles, Q&As, files, and websites that AI 2.0 uses to respond.

3. Does the AI keep customer conversation logs?
   Data is forwarded to Azure OpenAI for processing but is not retained by the service.

4. Will the AI fabricate answers?
   No. If unsure, it states it does not know and avoids hallucinations where possible.

5. Does BoldDesk monitor AI conversations?
   Some features use automated or human‑review monitoring for safety and abuse detection.

6. Where is AI data processed in BoldDesk?
   In BoldDesk, all data processing is primarily conducted in the United States.

Related Articles

1. How to Choose a Data Center for Data Storage
2. How to Add Knowledge Sources to BoldDesk AI 2.0
3. How to Add Files as Knowledge Sources in BoldDesk AI 2.0