Strategic Intelligence Copilot Safety Implementation Plan

Role: Chief Strategy Officer & Enterprise AI Risk Director

Context

• Business unit: {{enterprise_unit}}
• Strategic business objectives: {{strategic_priorities}}
• Proprietary data tiers: {{confidential_data_classes}}
• Corporate oversight framework: {{governance_framework}}
• Executive review frequency: {{decision_cadence}}
• Primary user tier: {{stakeholder_group}}

Task

Construct a safety and reliability implementation plan for an executive strategy copilot, ensuring analytical integrity, source attribution, and ironclad prevention of confidential data exfiltration.

Method

1. Define boundary parameters around {{strategic_priorities}} to focus copilot synthesis on approved corporate goals.
2. Establish zero-retention and redaction policies for all assets categorized under {{confidential_data_classes}}.
3. Architect citation-grounding requirements that reject market analysis lacking verifiable empirical source references.
4. Embed bias detection routines to identify over-optimistic revenue projections or ungrounded strategic assumptions.
5. Map data-flow boundaries ensuring {{enterprise_unit}} proprietary insights do not cross unauthorized departmental partitions.
6. Standardize explainability requirements tailored for consumption by {{stakeholder_group}}.
7. Align copilot decision-logging mechanisms with {{governance_framework}} audit requirements.
8. Establish scheduled review checkpoints synchronized with {{decision_cadence}}.

Constraints

• Copilot MUST NOT synthesize recommendations using uncited external data sources.
• Model MUST strictly block unredacted ingestion of {{confidential_data_classes}}.
• Strategy outputs must remain descriptive and analytical, never autonomously deciding resource allocation.
• System logs must preserve full audit trails without storing raw sensitive payload text.

Output format

Deliver an executive-level implementation plan structured into:

1. Strategic Safety Objectives & Scope (Executive summary)
2. Confidential Data Protection Architecture (Access & redaction controls)
3. Analytical Integrity & Grounding Rules (Factuality verification protocols)
4. Stakeholder Interface & Transparency Standards (Format for {{stakeholder_group}})
5. Governance Cadence & Audit Matrix (Alignment with {{decision_cadence}})

Self-review

• Verify that all data categories in {{confidential_data_classes}} have explicit isolation rules.
• Ensure alignment with the compliance standard named in {{governance_framework}}.
• Confirm that the output format directly addresses {{stakeholder_group}} workflows.