Skip to content

CSlawyer1985/legal-skillhub / skills/ai-compliance

ai-compliance

Verified

AI compliance analysis for EU AI Act, ISO 42001, NIST AI RMF, GDPR, OECD, financial services regulations (SEC, FCA, FINRA, DORA, MiFID II), and other frameworks. Use when asked to generate a compliance checklist for an AI tool or use case, determine if a risk assessment is required, score an AI tool's risk level, identify where an AI tool or use case could run afoul of regulatory requirements, perform a gap analysis, recommend remediation steps, assess a vendor, draft an acceptable use policy, map training requirements, or review jurisdiction-specific AI rules. Triggers on phrases like "compliance checklist", "risk assessment", "risk score", "does this comply", "EU AI Act", "ISO 42001", "NIST AI RMF", "AI governance", "gap analysis", "vendor assessment", "acceptable use policy", "MNPI", "AI training requirements", or any request to evaluate an AI tool or use case for regulatory compliance or risk.

references/vendor-assessment.md

View on GitHub
# Vendor AI Risk Assessment Questionnaire

Use this questionnaire when evaluating any third-party AI tool before procurement or continued use.

## Section 1 — Company & Product Overview
- [ ] Company name, HQ jurisdiction, parent company
- [ ] Product name and version assessed
- [ ] Is this a consumer or enterprise product?
- [ ] Enterprise tier available? (consumer use must be prohibited for business data)
- [ ] SOC 2 Type II report available? (request copy)
- [ ] ISO 27001 certified?
- [ ] ISO 42001 certified or in progress?
- [ ] Regulatory compliance claims (GDPR, CCPA, HIPAA, FedRAMP, etc.)

## Section 2 — Data Handling & Retention
- [ ] Are user prompts/queries stored? For how long?
- [ ] Is query data used to train models? (opt-out available?)
- [ ] Is query data shared with third parties or subprocessors?
- [ ] List of subprocessors available and reviewed?
- [ ] Data residency options available (EU, US, etc.)?
- [ ] Can data retention be disabled at enterprise tier?
- [ ] Data deletion process and SLA documented?
- [ ] Data portability/export available?

## Section 3 — Data Protection & Legal
- [ ] Data Processing Agreement (DPA) available?
- [ ] Standard Contractual Clauses (SCCs) available for EU data transfers?
- [ ] Privacy policy reviewed and acceptable?
- [ ] Does the vendor claim ownership of inputs or outputs?
- [ ] Intellectual property protections for firm content in outputs?
- [ ] Breach notification SLA (how quickly will they notify)? — 72 hours required for GDPR
- [ ] Right to audit contractually available?

## Section 4 — Security Controls
- [ ] Encryption in transit (TLS 1.2+)?
- [ ] Encryption at rest?
- [ ] Access controls and authentication options (SSO/SAML, MFA)?
- [ ] Role-based access control available?
- [ ] Vulnerability disclosure / bug bounty program in place?
- [ ] Penetration testing conducted? Frequency?
- [ ] Incident response plan documented?

## Section 5 — AI-Specific Risk
- [ ] Model training data sources documented?
- [ ] Model bias evaluation conducted and published?
- [ ] Hallucination/accuracy limitations documented?
- [ ] Content filtering / safety controls in place?
- [ ] Prompt injection mitigations in place?
- [ ] AI-generated output labeling available?
- [ ] Human oversight mechanisms supported?
- [ ] Explainability/interpretability features available?

## Section 6 — Operational Risk
- [ ] SLA and uptime guarantees acceptable?
- [ ] Vendor financial stability assessed (funding, public company)?
- [ ] Dependency risk — what happens if vendor shuts down?
- [ ] API stability and versioning policy?
- [ ] Geographic availability (outage risk for EU/US split ops)?

## Scoring Guide
Rate each section 1–5. Total score determines procurement path:
- **41–50**: Low risk — standard procurement
- **31–40**: Medium risk — require DPA + security review before approval
- **21–30**: High risk — require CISO approval + additional controls
- **<21**: Very high risk — block or require executive exception

## Common Findings by Tool
| Tool | Key Risks | Enterprise Tier |
|---|---|---|
| Perplexity | Query retention by default, US-based, consumer tier lacks DPA | Perplexity Enterprise Pro |
| ChatGPT | OpenAI trains on data unless opted out, SOC 2 available | ChatGPT Enterprise |
| Microsoft Copilot | M365 data residency, EU Data Boundary available | M365 Copilot (existing licensing) |
| Gemini | Google data policies, EU residency options | Google Workspace |
| Jasper.ai | Content stored, US-based | Jasper Business |
| Grammarly | Reads all text typed, wide data access | Grammarly Business |
receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementationStars: 323kUpdated Claude CodeCodexVerifiedInstall
requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirementsStars: 323kUpdated Claude CodeCodexVerifiedInstall
using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallbackStars: 323kUpdated Claude CodeCodexVerifiedInstall
verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions alwaysStars: 323kUpdated Claude CodeCodexVerifiedInstall