Information Classification Framework
Information Classification: Tier 2 - Internal
Authored by: Myron (5 May 2025)
Last modified by: Myron (31 May 2025)
Scope: All Mynt business units, staff, contractors and AI assistants handling company data
Keywords for AI/Search: data classification, confidentiality tiers, disclosure rules, AI access control, sensitive information, regulatory compliance, security policy
Summary
Mynt’s tier system labels every document according to sensitivity—from publicly shareable resources to strictly restricted files. Clear labelling protects client trust, supports regulatory compliance and lets our AI assistants answer queries without leaking confidential sources.
Purpose
Standardise how all content is tagged so the right people (and only the right people) see it.
Guide staff and AI tools on whether, and how, information may be quoted, summarised, or withheld.
Mitigate risk of accidental disclosure of strategic, personal or financial data.
Enable automation by giving RAG systems an unambiguous flag that filters responses before they reach end-users.
Why it matters
Protects client trust and regulatory compliance.
Prevents accidental disclosure of strategic or personal data.
Enables our AI tools to filter answers automatically.
The Tiers
Tier
Label
Audience
External sharing?
Comments
1
Open
Public
Semi Permitted
AI assistants may use this information to answer question from external parties. However, it should never share the source information.
2
Internal
All Mynt Staff
Prohibited
Mainly used by AI assistants to help staff with work such as lesson plans, or information about company policies.
3
Confidential
Managers
Prohibited
Confidential information such as salary data, strategic policies, and others.
4
Restricted
C-suite, Board Members, Shareholders, and Designated Individuals
Prohibited
Highly sensitive data such as financial reports and strategic documents.
How to tag content
Insert the label "Information Classification" and one of the three tiers, Tier 1 - Open
, Tier 2 - Internal
or Tier 3 - Confidential
, at the top of every document or article.
Guidance for AI systems
The AI may freely use the Tier‑1 information for responses. It can summarize, or answer with context, but never share the source information.
For Tier‑2 content, the AI must confirm the user is a current employee before disclosure.
The AI must refuse to reveal Tier‑3 or higher content unless the user is a verified manager.
Last updated