What this sets up
People leaders produce a distinctive artefact set: decision memos, team communications and one-on-one preparation. Microsoft 365 Copilot has no persistent project-instructions field, so this configuration gives you a standing instruction block to paste at the start of a working session (or save as a reusable prompt where your tenant offers prompt saving) and a small file pack kept in one OneDrive or SharePoint folder that Copilot grounds on when you reference or attach the files. The payoff is consistency: the same decision-memo skeleton every time makes decisions comparable months later, and a comms rhythm file stops the weekly note drifting in tone. The method comes from the AI Setup module LM-S01; this page packages the leaders playbook for Copilot specifically.
What never goes in
Nothing about an individual's performance, health or personal circumstances enters this set-up, full stop. Team members are referenced by initials or role only. That rule is older than the platform, but Microsoft's documentation makes it concrete:
- Your prompts are a discoverable record: stored Copilot interactions include "the user's prompt and Copilot's response, including citations", held in a hidden folder of your mailbox that "compliance administrators can search with eDiscovery tools". Microsoft warns that "Messages visible in your AI apps are not an accurate reflection of whether they are retained or permanently deleted", and a departed employee's Copilot history can persist and stay searchable. A prompt speculating about a team member is a discoverable document with your name on it; write every prompt as if it were a work email.
- Microsoft itself draws a line here: its documentation states that in certain Microsoft 365 Copilot scenarios it restricts generative AI "inferences, judgments, or evaluations about an employee's performance, attitude, internal or emotional state, or personal characteristics". Do not ask for what the platform is designed to refuse; assessments of people are yours to make, outside the tool.
- Work only ever happens signed in with your Microsoft Entra work account, with the enterprise data protection shield showing; the consumer Copilot app on a personal account is a different product with different terms. Under enterprise data protection, "Prompts, responses, and data accessed through Microsoft Graph aren't used to train foundation LLMs, including those used by Microsoft 365 Copilot."
- Copilot "only surfaces organizational data to which individual users have at least view permissions". As a leader you hold wide permissions: Copilot can ground on material your team cannot see, so check any Copilot-drafted communication for content your audience is not cleared to read before sending.
- Web grounding leaves the tenant: when Copilot queries Bing, Microsoft is "an independent data controller" for those queries, outside the Data Protection Addendum, and short prompts can reach Bing largely intact. Keep names, restructure plans and anything commercially sensitive out of web-grounded prompts.
Sources, all verified 14 August 2026: Data, Privacy, and Security for Microsoft 365 Copilot, Enterprise data protection in Microsoft 365 Copilot, Retention policies for Copilot, Copilot privacy and protections, Copilot FAQ.
The instruction text
Start each working session by pasting this as your first message, adjusting the folder name to yours:
The files to attach
Write these five markdown files, keep them in one OneDrive or SharePoint folder, and reference or attach them as each task needs. Each links to a published artefact to build it from:
- writing-style.md: your voice profile, built with the voice-profile interview in module LM-S01, Part 1.
- team-context.md: what the team does, current priorities, standing constraints; no personal detail about members.
- decision-memo-template.md: your decision-record structure. The AI meeting decision record article and the decision record recipe show working shapes, the recipe downloadable as raw markdown.
- comms-rhythm.md: what you send, to whom, how often, in what format.
- achievements-log.md: a running record of the team's wins, updated weekly, entry format per module LM-S02, Part 1; it quietly becomes the best input to recognition, reporting and your own performance review.
The first three prompts
Governance guardrails
- The people line is bright: one-on-one preparation produces questions to ask, never conclusions about the person, and anything that reads as an assessment of an individual moves out of the tool and into your normal management process.
- Decisions stay yours: the memo records reasoning for a decision you have made or are making; drafting the memo is not making the decision, and the risks-accepted line is written by you, not accepted on your behalf.
- Read every Copilot-grounded draft for material your audience cannot see before sending; your permissions are wider than theirs.
- Keep the prompt history clean enough to stand in a file review: professional, de-identified, no speculation about people.
- Check your organisation's Copilot rollout guidance first; tenant restrictions on web grounding, agents or file attachment are part of your control environment.
About this configuration
Built against: Microsoft 365 Copilot Chat with a Microsoft Entra work account, August 2026. Platform behaviour, licensing and terms change; re-check the sources above before relying on any data-handling claim. TheAICommand is independent: no platform vendor paid for, reviewed or influenced this configuration, and we accept no payment for coverage. See ownership and funding.
TheAICommand. Intelligence, At Your Command.