What the AI can and can't see
The data boundaries behind a connected AI assistant — which Workspaces it can reach, what it can read versus change, and what it never sees, all set by you.
Updated 23 June 2026
When you connect an AI assistant to In Parallel over MCP, what it can reach is bounded by choices you make — which Workspaces you share, which actions you approve, and whether people’s names are masked.
Workspace hierarchy affects your own access: ancestor members can reach descendant Workspaces. An AI client still needs its separately authorized Workspace access; choosing a preferred assistant for Send to AI does not create that grant.
What it is
An AI assistant connected over MCP only ever sees In Parallel through the access you grant it. There is no hidden, broader view: the assistant reaches your data through a defined set of tools, and three layers of control sit in front of those tools — the Workspaces you shared at connect time, the per-tool permissions you set, and the optional masking of people. This article explains each boundary so you know exactly what a connected assistant can and can’t reach before you point it at sensitive work.
Boundary 1 — only the Workspaces you shared
When you connect a tool, In Parallel’s consent screen asks which Workspaces to grant access to:
- Full access shares all your current and future Workspaces.
- Or you tick individual Workspaces to share only those.
The assistant can only see Workspaces in that grant. A Workspace you didn’t share is invisible to it — it won’t appear in list_workspaces, and no tool can reach into it. To widen or narrow the grant later, reconnect and choose differently. (The grant is also bounded by your own access — connecting an assistant never lets it see more than you can.)
Boundary 2 — reading is free, changing needs approval
Within the Workspaces you shared, tools split into two kinds:
- Read tools look things up — meetings, transcripts, decisions, action items, goals, risks, open questions, and any health tools shown in the live tool list — and never change anything. These run freely.
- Write tools create or change records. In clients that support per-tool permissions (like Claude), write and delete tools default to Needs approval, so the assistant has to ask before it acts. You can tighten this further — switch any tool or group to Blocked, or set a Custom rule.
A handful of write tools are irreversible — deleting a Workspace, publishing or rejecting a changelog entry — and stay on Needs approval for good reason. You set all of this when you connect; see step 6 of the Claude walkthrough in Connect your AI tool.
Some write tools are also ownership-gated: the changelog tools, for example, only reach entries in Workspaces you own, even if you can read others.
Boundary 3 — names are visible unless you mask them
By default, a connected assistant can receive real names and email addresses in the data it reads. A Workspace Owner can turn on MCP privacy: In Parallel then replaces directory-known people with stable pseudonyms, such as Person 4F2A1C, in MCP responses.
Masking is deliberately narrow. It covers people In Parallel knows, such as Workspace members and meeting participants, but filenames, URLs, object names, and unrecognized names in free text can remain visible. Review the source data before using MCP with sensitive material.
What it can and can’t reach — at a glance
| The assistant can… | The assistant can’t… |
|---|---|
| Read meetings, transcripts, decisions, action items, goals, risks, open questions, and observations — in the Workspaces you shared | Reach a Workspace you didn’t grant access to |
| Create or change records when the client permission allows it | Bypass the Workspace grant or client permissions you configured |
| Act on changelog entries in Workspaces you own | Publish or reject changelog entries in Workspaces you don’t own |
| Receive known people as pseudonyms when Company MCP privacy is on | Treat masking as complete redaction of filenames, URLs, or arbitrary free text |
| Use the current set of tools your client fetches on connect | Reach anything outside the In Parallel tools — there’s no back door to your data |
A note on the AI built into In Parallel
This article is about an external assistant you connect over MCP. The AI built into In Parallel applies different controls to different artifacts. Maintained narrative pages can refresh automatically and show freshness, source, and version cues. Structured execution changes and external actions use the approval controls available for that workflow. In every case, AI-generated wording is not proof of a decision or authorization. See AI in In Parallel.
Related
- What you can do with In Parallel’s AI — the use-cases these boundaries make safe.
- MCP privacy: mask people in AI responses — hide names and emails from your assistant.
- Connect your AI tool to In Parallel — where you set the Workspace grant and per-tool permissions.
- In Parallel MCP tools reference — every tool, and which ones read versus write.
- Governance, security, and trust — In Parallel’s access controls and data handling.
- AI in In Parallel — the in-product AI layer and its review gate.