Secure Microsoft Teams for AI
Microsoft Graph's Channel.ReadBasic and Chat.Read permissions hand AI access to every team, every channel, and every 1:1 chat across your tenant. PortEden sits between your AI agent and Graph, enforcing team and channel-level rules and redacting sensitive content before it reaches the model.
How PortEden Protects You
Six layers of security between AI and your data.
Team and Channel Allowlists
Permit AI in specific teams (Engineering, Product) and block sensitive ones (Leadership, HR, Finance). Rules apply to message history, channel searches, and live lookups.
Private Chat Blocking
Hide every 1:1 chat and group chat from AI by default. Whitelist specific threads if you opt in. AI never sees private conversations or HR escalations.
Content Redaction
Mask credentials, customer PII, payment data, and access tokens pasted into Teams channels before AI reads them — including content in attached cards and adaptive messages.
Action Limits
Restrict whether AI can post messages, react, or create channels. Read-only mode is the default; write actions require explicit per-channel grants.
Time Windows
Limit AI to recent messages only — e.g. last 24 hours or last 7 days — so old project chatter and stale escalations stay out of context.
Full Audit Trail
Every channel read, search, and blocked request is logged with the agent identity, timestamp, and rule that fired. Trace any incident back to the exact request.
Get Started in 3 Steps
Install — CLI or MCP
Install the PortEden CLI or add the cloud MCP connector to your AI client, then authorize PortEden in your Microsoft 365 tenant.
Set Team and Channel Rules
Allowlist teams and channels for AI access, block 1:1 and group chats, and enable content redaction patterns.
Connect Your Agent
Point your AI agent at PortEden instead of Microsoft Graph. Every read, search, and post is filtered, redacted, and logged.
Without vs. With PortEden
Without PortEden
- Channel.ReadBasic and Chat.Read grant access to every team, channel, and chat
- 1:1 chats and group chats with HR, legal, and exec conversations fully exposed
- Pasted credentials, API keys, and PII flow straight into the AI model
- No granular audit of which teams or messages AI accessed
- Revoking AI requires removing the Graph app from your tenant
With PortEden
- Team and channel allowlist limits AI to the surfaces you explicitly permit
- 1:1 and group chats blocked by default; opt-in per conversation
- Credential and PII redaction strips sensitive content before AI sees it
- Read-only mode prevents AI from posting, reacting, or creating channels
- Full per-agent audit log of every read, search, and blocked request