SoftBank AI Ethics Policy

[Note]

• *
  Please note that this document is an unofficial translation and was prepared for reference purpose only. The original release is in Japanese.

SOFTBANK CORP. has consistently contributed to society through the information revolution, based on its management philosophy of “Information Revolution — Happiness for everyone”. Technology should be used only for the happiness of people, and should never cause them grief or suffering.
We believe that AI, as a rapidly advancing and diversifying technology created by humankind, has the potential to bring creative destruction to all industries more widely than any other technology in history, and to contribute more than ever to realize the human happiness. As a corporate group that believes in the potential of AI and is a leader in its development, we aim to comply with laws and social norms, and create a world in which AI is properly used for the happiness of people.

This policy is intended to complement on AI ethics, but not contradict to the following previously established policies.
SoftBank Human Rights Policy
SoftBank Privacy Policy
Information Security Policy

1. 1.
   Human-Centered Principles
   We believe that AI exists to make people happy and should be used only as a decision-making aid for people who use it to solve social problems. Therefore, humans should be appropriately involved in each stage of AI design, development, provision, and use. We aim to create AI that makes people happier through its use.
2. 2.
   Respect for Fairness
   Fair AI cannot be realized unless we humans have fair values. We respect all people regardless of race, nationality, age, religion, gender, etc. We will design, develop, provide, and use AI based on the same principle, paying attention to the existence of social minorities, such as racial minorities, and sexual minorities, which are not readily apparent in commonly distributed data. We aim to realize an inclusive society through technology.
3. 3.
   Pursuit of Transparency and Accountability
   In order to coexist and develop together with AI, it is important to create an environment in which people can understand the results of AI-based judgments and verify the basis of those judgments when necessary. Therefore, we will review AI and manage records/evidence under appropriate governance and clear responsibility frameworks, and we will strive for transparency in explaining how we use the results of AI judgments. We aim to design, develop, provide, and use AI with verifiability regarding the reliability of the AI and a high level of accountability regarding the basis for the judgments.
4. 4.
   Safety Assurance
   We believe that AI that contributes to people's happiness can only be realized with a high level of safety. We will assess the risks of AI, restrict its use as necessary, and continue monitoring and evaluating AI even after operation. While doing so, we shall strive to design, develop, provide, and use AI that does not threaten the life, liberty, dignity, or property of individuals, and to create a world in which all people can use AI safely.
5. 5.
   Privacy Protection and Security Assurance
   Data is indispensable for the design, development, and provision, and use of AI. Without a large amount of accurate data to educate AI, its development and contribution to society cannot be realized. That is why we will appropriately manage and operate various types of data, including personal data, in accordance with internal rules, relevant laws and regulations, and social ethics to prevent unauthorized access, etc., and protect privacy so that people can use AI with peace of mind. Also, we will respond promptly in the event of an incident.
6. 6.
   Education for AI talent and Literacy
   We have been focusing on the development of advanced technology. While technology is advancing by leaps and bounds, the people who utilize it must also be constantly evolving and capable of handling technology responsibly. In order to realize a society in which everyone can enjoy the benefits of AI to the fullest, while taking into account the evolution and new forms of AI utilization, so that people can understand the characteristics, limitations, and risks of AI, we will actively engage in education and training to enhance AI-related skills and literacy.
7. 7.
   Promotion of Innovation
   We will actively incorporate technical trends, case studies, and professional knowledge regarding AI through collaboration with industry, academia, and government, and strive to secure interconnectivity and interoperability between AI systems and services. Also, we will establish mechanisms that allow data and technology to be appropriately utilized across organizations and domains, and promote innovation that contributes to the resolution of social problems and the creation of new value.

Revision of this Policy
We will review this policy as necessary and respond flexibly in light of national and regional guidelines on AI, changes in people's lifestyles and environments, accumulated examples in the industry, developments in technology itself, and dialogues with various stakeholders. We will flexibly respond to such changes as necessary.