Español

Ethical Best Practices Manual for the Use of Artificial Intelligence

Introduction

Holy Mountain is committed to the ethical and responsible implementation of artificial intelligence (AI) technologies, strictly aligning with the European AI Regulation. Our mission is to merge human creativity with the generative capacity of AI, respecting intellectual property rights and ensuring compliance with current European regulations. This manual outlines the ethical and legal commitments of Holy Mountain at every stage of AI development and use.

1. Commitment to Intellectual Property Rights

Holy Mountain recognizes the importance of intellectual property (IP) rights in the creation and use of content. In compliance with the European AI Regulation, we commit to:

• Respect and Protection of Copyrights: We will not use copyrighted content without the appropriate authorization from the rights holders. Every resource used will be carefully reviewed to ensure compliance with current regulations.

• Prior Authorization: Before using protected content for the training of customized models or specific projects, we will obtain the necessary authorizations from the rights holders. This includes clear licensing agreements for any creative resource utilized.

• Transparency in Content Use: Holy Mountain is committed to providing precise information about the origin of data used in its creative processes. This includes detailing data sources and, when necessary, communicating any limitations related to their traceability.

Use of Foundational Models

Holy Mountain utilizes foundational generative AI models, such as Stable Diffusion, MidJourney, Flux, among others, to develop innovative creative content. While we use these tools under license and in compliance with their terms of use, we recognize that we do not control the original datasets used for their initial training.

To ensure transparency and maintain high ethical standards, we adopt the following internal measures concerning Foundational Models:

  1. Responsible Selection: Holy Mountain evaluates the licenses and terms of use for each tool to ensure its application is ethical and legal.

  2. Customized Adaptation: Whenever possible, we train derivative models (LoRA, embeddings, etc.) using data provided or licensed directly by our clients, ensuring that the generated content reflects their values and complies with regulations.

  3. Internal Communication: We internally document the use of these models and any associated limitations to be prepared for any inquiries or requests for information.

  4. Client Transparency: We inform our clients about the AI tools used in their projects, explaining their capabilities and relevant limitations without the need to embed disclaimers in the generated visual assets.

  5. Claims Management: If a copyright holder raises a potential misuse related to foundational models, we offer client support to address the issue with the corresponding technology provider.

2. "Human First" Philosophy and Responsible Use of AI

At Holy Mountain, we firmly believe in the primacy of human talent in the creative process. Our "Human First" philosophy ensures that all AI tools employed are always under the direction and control of our artists and creatives. We commit to:

• AI as a Support Tool: We use AI as a tool to enhance and complement human creativity, never as a substitute. Ideas, concepts, and creative decisions are driven by our professionals.

• Constant Human Oversight: We ensure that every stage of the process, from conceptualization to final delivery, is guided and directed by human supervision.

• Ethics and Responsibility: We ensure that AI use aligns with our ethical values and those of our clients, avoiding any outcomes that could be harmful, inappropriate, or contrary to our principles.

• Client Transparency: We clearly communicate how AI is integrated into our processes, ensuring that our clients understand its role as a support tool and the added value it brings to our services.

3. Protection of Fundamental Rights and Compliance with the General Data Protection Regulation (GDPR)

Holy Mountain prioritizes the respect of individuals' fundamental rights in all AI-related activities:

• Non-Discrimination: Our AI systems will not generate discriminatory or harmful content toward any group. We implement practices to prevent bias and promote inclusion across all projects.

• Security and Privacy: We handle all data used in our processes with the utmost care, in compliance with European regulations such as the General Data Protection Regulation (GDPR).

4. Basic Documentation and Risk Consideration

Holy Mountain maintains a simple record of the AI tools used and the data and content sources employed in projects. We commit to:

• Tool and Source Registry: Keeping an up-to-date list of the AI tools and content sources used, ensuring traceability and compliance with regulations.

• Risk Consideration: Identifying and mitigating potential risks associated with the use of AI in our projects, particularly concerning intellectual property rights and the generation of inappropriate content.

5. Collaboration and Responsibility with Clients

Holy Mountain is committed to maintaining a transparent and trustworthy relationship with our clients by providing clear information about the AI systems used:

• Transparent Communication: We will inform our clients about any relevant considerations related to the use of our AI services, including technical limitations and intellectual property aspects.

• General Guidance: We will offer information on how AI can be integrated into their projects, helping them understand the capabilities and limitations of the generative technologies used.

• Regulatory Compliance: We will ensure that our AI solutions fully comply with current European regulations and share relevant information to keep clients informed.

• Confidentiality and Security: We will protect the information and data provided by our clients, guaranteeing their confidentiality and security at all times.

6. Continuous Training and Adaptive Process Improvement

Holy Mountain maintains a commitment to continuous training and improvement, tailored to the evolving AI environment and the specific needs of our clients. This approach translates into:

• Training in Relevant AI and Regulations: We participate in accessible and specific training activities to gain a proper understanding of technical, legal, and ethical advancements in AI.

• Critical Evaluation of New Tools: We examine new tools and technologies that may enhance our creative capabilities, ensuring they are safe, ethical, and aligned with our needs and goals.

• Specific Adjustments and Improvements: We continuously review and adapt our practices to stay aligned with ethical values, market demands, and client needs.

• Development of Creative and Technical Skills: We foster continuous growth of skills related to creativity and AI to ensure our projects reflect the highest quality.

• Feedback and Learning: We incorporate lessons learned from each project to improve our practices and future outcomes.

Conclusion

Holy Mountain reaffirms its commitment to developing and applying AI technologies ethically and responsibly, in alignment with the European AI Regulation. This manual reflects our values and ensures that our actions uphold the highest standards of integrity, transparency, and responsibility in the creative sector.