What problem does it solve? Legal and compliance teams need a documented, defensible record of how each AI system was evaluated before deployment, but writing an AI impact assessment from scratch is slow and inconsistent across reviewers. ## Core Features & Use Cases - Structured intake and risk triage: Runs a conversational intake covering the system, affected parties, data inputs, oversight, accuracy, and deployment stage, then selects a fast-track or full assessment based on governance tier. - Regulatory classification per regime: Researches each regime in the configured regulatory footprint (e.g., EU AI Act, GDPR, NYC Local Law 144, Colorado AI Act), classifies the system's risk tier with pinpoint citations, flags prohibited-practice exposure, and identifies whether a separate fundamental-rights impact assessment is required. - Policy diff and recommendation: Cross-checks the use case against the organization's AI policy commitments, produces a risk-and-mitigation table, and outputs an approval recommendation with conditions and handoffs (PIA, vendor review, regulatory gap analysis). - Use Case: A company wants to deploy an AI résumé screening tool for HR. The skill intakes the system details, classifies it under applicable employment and AI regimes, flags bias-testing gaps, and produces a house-style assessment with deployment conditions. ## Quick Start Ask the assistant to run an AI impact assessment for your use case, for example: run an AIA for AI résumé screening used by our HR team.
