Hook Habitat Limited (trading as Karen Howley)
Last updated: July 2025
Introduction
This is a concise overview of our AI Policy. A full version, including detailed procedures and controls, is available on request for clients and partners. Please email karen@karenhowley.com.
Our commitment
We use AI to enhance – not replace – human expertise. All AI activities comply with GDPR and the forthcoming EU AI Act (Article 50 transparency).
Core principles
Human Oversight: Every AI-assisted output is reviewed and approved by Karen Howley.
Data Responsibility: We process only data covered by our Privacy Policy, using enterprise-grade platforms for client information.
Transparency: Any AI-assisted content is clearly labelled.
Bias & Fairness: We champion critical thinking. Karen applies her inclusion and diversity training and, where needed, prompts AI to consider multiple perspectives.
AI Policy
Approved tools
Managed Platforms (client data OK): Microsoft 365 Copilot; ChatGPT Team (Enterprise)
Individually Managed (restricted use): Canva AI (paid subscription with MFA; client decks retained up to 6 years)
Experimentation Only (no client data): Eleven Labs; Perplexity; Claude; Gemini, Midjourney
What we won’t do
No fully automated decisions affecting people without a human sign-off
No uploading of sensitive client data to public AI services
No deployment of AI outputs externally without final human validation
Security & retention
All AI outputs are stored securely in Microsoft 365.
Meeting transcripts auto-delete after 120 days; Canva designs retained up to 6 years, then purged.
Keeping it up to date
We review and update this summary annually or whenever we adopt significant new AI tools or practices.