AI Transparency Policy

Last updated: 5 April 2026

Our commitment to transparency

Mexzungu Group is an AI-augmented Studio. We use artificial intelligence as a core part of our operations, not as a side tool. We believe you should know exactly how.

Janis: AI Chief of Staff

Janis is our operational partner, powered by Anthropic's Claude. She handles:

• Email triage and response drafting
• Document research and preparation
• Calendar management and scheduling
• Financial tracking and expense filing
• Client communication support (substantive messages are always reviewed by Pepe Carrillo before sending)

Janis is not a chatbot. She is an integrated operational system that manages day-to-day business processes across the Studio.

The Rock Council: Virtual C-Suite

We use AI-powered personas to bring structured, multi-perspective analysis to every major decision:

• Lars (COO): Operations and delivery tracking
• Freddie (CMO): Brand, content, and audience growth
• Gene (CFO): Finance, cash management, and tax compliance
• Bowie (CTO): Technology infrastructure and reliability
• Bono (CGO): Growth, pipeline, and network activation
• Lemmy (CLO): Legal, compliance, and risk management
• Bruce (CSO): Strategy, competitive intelligence, and long-range planning

These personas provide analytical frameworks and recommendations. All final decisions are made by Pepe Carrillo.

Our AI principles

1. Human authority. All substantive client decisions, contracts, pricing, legal positions, and public communications require explicit approval from Pepe Carrillo before execution.
2. Transparency. We disclose our use of AI to clients, partners, and visitors. This page is part of that commitment.
3. Data protection. Client data processed through AI systems is handled per our Privacy Policy. Our AI provider (Anthropic) does not use API data for model training.
4. Quality control. AI-generated content and communications are reviewed before delivery. We do not send unreviewed AI output to clients on substantive matters.
5. Right to human interaction. Any client, partner, or visitor can request direct, human-only interaction at any time by contacting pepe@mexzungu.com.

Technology stack

• AI Model: Anthropic Claude (Opus and Sonnet)
• Infrastructure: Self-hosted on European servers (Hetzner, Helsinki, Finland)
• Data residency: Processing infrastructure within the European Union
• Provider policy: Anthropic does not train on API data (per their commercial data policy)

EU AI Act compliance

In accordance with the EU AI Act (Regulation 2024/1689), we disclose that AI systems are used in our business communications and service delivery. Our AI systems are classified as limited-risk under the Act, and this transparency page fulfills the disclosure obligations under Article 50.

Changes to this policy

We will update this page when our AI practices change materially. The "last updated" date at the top reflects the most recent revision.

Contact

Questions about our AI practices: pepe@mexzungu.com