Japan – AI guidelines for business 2024

The AI Guidelines for Business, released by the Ministry of Internal Affairs and Communications and the Ministry of Economy, Trade and Industry published on April 19, 2024, provide a comprehensive framework for the development, provision, and use of Artificial Intelligence (AI) in business settings. These guidelines aim to promote innovation while mitigating risks associated with AI technologies, focusing on a human-centric approach. 

Contents

  1.  Preface
  2.  Definitions
  3. Society to Aim for with AI and Matters Each AI Business Actor Works On.
  4. Basic Philosophies
  5. Principles
  6. Common Guiding Principles
  7. Common Guiding Principles for AI Business Actors Involved in Advanced AI Systems
  8. Building AI Governance
  9. Matters Related to AI Developers
  10. Matters Related to AI Providers
  11. Matters Related to AI Business Users

Preface

The introduction highlights the rapid evolution of AI technologies and their increasing application in various sectors. It emphasizes Japan’s commitment to Society 5.0, a human-centric approach that aims to achieve both economic growth and solutions to social challenges through advanced integration of cyberspace and physical space. The guidelines also address the emerging risks associated with AI, particularly from generative AI, and stress the importance of developing a governance framework to manage these risks effectively.

Society Goals and Philosophies:

Basic Philosophies

  1. Dignity: Respect for human dignity, ensuring that AI enhances rather than diminishes human capabilities.
  2. Diversity & Inclusion: Creating a society where diverse backgrounds contribute to collective well-being.
  3. Sustainability: Leveraging AI to address global challenges, including social disparities and environmental issues

Principles and Common Guiding Principles

The document further categorizes guiding principles into actionable items for AI business actors, emphasizing cooperation with societal stakeholders to enhance the positive impacts of AI while minimizing risks.

It outlines principles for AI business actors, emphasizing the need for compliance with human-centric values, safety, fairness, privacy protection, security, transparency, and accountability. The section also highlights the importance of collaboration among various stakeholders.

Common Guiding Principles

The guidelines provide a framework for AI actors based on:

  1. Human-Centric Approach: Respect for human rights and autonomy.
  2. Safety: Ensuring the safety of AI systems.
  3. Fairness: Addressing biases in AI systems.
  4. Privacy Protection: Safeguarding personal data.
  5. Security: Implementing security measures throughout the AI lifecycle.
  6. Transparency: Ensuring stakeholders have access to relevant information.
  7. Accountability: Each actor must take responsibility for their actions.

Matters Related to AI Developers

Part 3 of this document outlines the fact that AI developers play a crucial role in shaping AI outputs and must carefully consider the ethical implications of their work. Key responsibilities include:

  1. Implementing proper data training and ensuring data quality.
  2. Developing AI systems that prioritize safety and human rights.
  3. Regularly evaluating AI models for bias and ensuring transparency in AI operations

Part 3 outlines the matters relating to AI developers and the actions they must take. The key points derived from this section are outlined below.

Data Pre-processing and Training:

  1. Proper Data Training: Collect training data through privacy-by-design and ensure proper handling of personal data in compliance with laws and regulations.
  2. Consideration for Bias in Data: Control the quality of the data to avoid biases in AI models and ensure that training data is properly represented.

When Developing AI:

  1. Development Considerations: Set clear policies regarding safe AI use to avoid unexpected harm and ensure proper performance under various circumstances.
  2. Consideration for Bias in Algorithms: Address potential biases introduced by technical elements making up AI models.

Deployment and Maintenance:

  1. Security Measures: Implement appropriate security measures throughout the development of AI systems.
  2. Ensuring Verifiability: Maintain records for follow-up verification and take measures to improve AI quality.

Post-Development Actions:

  1. Consideration of Latest Trends: Stay updated on new attack methods targeting AI systems.
  2. Providing Relevant Information: Share timely information about AI systems with stakeholders.

Accountability:

  1. Explanation of Conformity: AI developers must communicate to AI providers about potential changes in AI performance and risks associated with it.
  2. Documentation: Prepare documents on AI development processes to improve transparency and accountability.

Contribution to Innovation:

  1. Engage in research and development efforts to advance the quality and reliability of AI while promoting collaboration among various sectors.
  2. These points highlight the responsibilities and considerations AI developers must focus on to ensure the safe and effective use of AI systems.

Matters Related to AI Providers

AI providers are tasked with ensuring that AI systems are used properly and ethically. Their responsibilities encompass:

  1. Implementing security measures to protect stakeholder interests.
  2. Providing clear documentation on AI systems to enhance traceability and transparency.
  3. Establishing protocols to prevent the misuse of AI systems.

Matters Related to AI Business Users

Part 4 of the document outlines the matters relating to business providers.

AI business users must utilize AI systems responsibly to derive maximum benefits while safeguarding ethical standards. Their duties include:

  1. Ensuring proper use of AI systems within the parameters set by providers.
  2. Actively mitigating risks associated with privacy violations and bias in AI outputs.
  3. Maintaining open communication with stakeholders regarding the use and implications of AI.

When Implementing an AI System:

  1. Actions against risks: Take measures to prevent AI from causing harm to the lives, bodies, properties, and minds of stakeholders, as well as the environment. Ensure proper performance under expected usage conditions and minimize risks from unexpected situations.
  2. Provision contributing to proper use: Establish correct considerations for using AI systems and services, use AI within the expected scope set by AI developers, and guarantee the accuracy and recency of training data.
  3. Consideration for bias in configurations and data: Ensure fairness of data at the time of provision and examine bias in referenced information.
  4. Deployment of mechanisms for protecting privacy: Implement privacy protection measures throughout the AI system’s life cycle.
  5. Deployment of mechanisms for security measures: Take appropriate security measures based on adopted technologies.
  6. Documentation of system architectures: Prepare documents describing the system architecture and data processing to improve traceability and transparency.

After an AI System or Service Starts to be Provided:

  1. Provision contributing to proper use: Periodically verify that the AI system is used for proper purposes.
  2. Countermeasures against privacy violation: Collect information on privacy protections and discuss strategies to prevent violations.
  3. Handling of vulnerabilities: Identify new risks and trends in vulnerabilities related to AI systems.
  4. Providing relevant stakeholders with information: Communicate important information about the AI system or service in a timely manner.
  5. Explanation to AI business users: Encourage proper use by explaining necessary considerations and precautions.

Conclusion

The AI Guidelines for Business serve as a vital resource for navigating the complexities of AI integration in business environments. By adhering to these guidelines, businesses can foster innovation while ensuring ethical practices and societal well-being. The guidelines are designed to evolve with the rapidly changing landscape of AI technology, ensuring they remain relevant and effective in promoting a balanced approach to AI development and usage.

In summary, the document outlines a comprehensive approach to AI governance in Japan, focusing on ethical considerations and collaborative efforts among all AI business actors to ensure the safe and beneficial use of AI technologies.