· 17) Non subversion

The power conferred by control of highly advanced AI systems should respect and improve, rather than subvert, the social and civic processes on which the health of society depends.
Principle: Asilomar AI Principles, Jan 3-8, 2017

Published by Future of Life Institute (FLI), Beneficial AI 2017

Related Principles

I. Human agency and oversight

AI systems should support individuals in making better, more informed choices in accordance with their goals. They should act as enablers to a flourishing and equitable society by supporting human agency and fundamental rights, and not decrease, limit or misguide human autonomy. The overall wellbeing of the user should be central to the system's functionality. Human oversight helps ensuring that an AI system does not undermine human autonomy or causes other adverse effects. Depending on the specific AI based system and its application area, the appropriate degrees of control measures, including the adaptability, accuracy and explainability of AI based systems, should be ensured. Oversight may be achieved through governance mechanisms such as ensuring a human in the loop, human on the loop, or human in command approach. It must be ensured that public authorities have the ability to exercise their oversight powers in line with their mandates. All other things being equal, the less oversight a human can exercise over an AI system, the more extensive testing and stricter governance is required.

Published by European Commission in Key requirements for trustworthy AI, Apr 8, 2019

· 1. The Principle of Beneficence: “Do Good”

AI systems should be designed and developed to improve individual and collective wellbeing. AI systems can do so by generating prosperity, value creation and wealth maximization and sustainability. At the same time, beneficent AI systems can contribute to wellbeing by seeking achievement of a fair, inclusive and peaceful society, by helping to increase citizen’s mental autonomy, with equal distribution of economic, social and political opportunity. AI systems can be a force for collective good when deployed towards objectives like: the protection of democratic process and rule of law; the provision of common goods and services at low cost and high quality; data literacy and representativeness; damage mitigation and trust optimization towards users; achievement of the UN Sustainable Development Goals or sustainability understood more broadly, according to the pillars of economic development, social equity, and environmental protection. In other words, AI can be a tool to bring more good into the world and or to help with the world’s greatest challenges.

Published by The European Commission’s High-Level Expert Group on Artificial Intelligence in Draft Ethics Guidelines for Trustworthy AI, Dec 18, 2018

· 4. Governance of AI Autonomy (Human oversight)

The correct approach to assuring properties such as safety, accuracy, adaptability, privacy, explicability, compliance with the rule of law and ethical conformity heavily depends on specific details of the AI system, its area of application, its level of impact on individuals, communities or society and its level of autonomy. The level of autonomy results from the use case and the degree of sophistication needed for a task. All other things being equal, the greater degree of autonomy that is given to an AI system, the more extensive testing and stricter governance is required. It must be ensured that AI systems continue to behave as intended when feedback signals become sparser. Depending on the area of application and or the level of impact on individuals, communities or society of the AI system, different levels or instances of governance (incl. human oversight) will be necessary. This is relevant for a large number of AI applications, and more particularly for the use of AI to suggest or take decisions concerning individuals or communities (algorithmic decision support). Good governance of AI autonomy in this respect includes for instance more or earlier human intervention depending on the level of societal impact of the AI system. This also includes the predicament that a user of an AI system, particularly in a work or decision making environment, is allowed to deviate from a path or decision chosen or recommended by the AI system.

Published by The European Commission’s High-Level Expert Group on Artificial Intelligence in Draft Ethics Guidelines for Trustworthy AI, Dec 18, 2018

Chapter 1. General Principles

  1. This set of norms aims to integrate ethics into the entire life cycle of AI, to promote fairness, justice, harmony, safety and security, and to avoid issues such as prejudice, discrimination, privacy and information leakage.   2. This set of norms applies to natural persons, legal persons, and other related organizations engaged in related activities such as management, research and development, supply, and use of AI. (1) The management activities mainly refer to strategic planning, formulation and implementation of policies, laws, regulations, and technical standards, resource allocation, supervision and inspection, etc. (2) The research and development activities mainly refer to scientific research, technology development, product development, etc. related to AI. (3) The supply activities mainly refer to the production, operation, and sales of AI products and services. (4) The use activities mainly refer to the procurement, consumption, and manipulation of AI products and services.   3. Various activities of AI shall abide by the following fundamental ethical norms. (1) Enhancing the well being of humankind. Adhere to the people oriented vision, abide by the common values of humankind, respect human rights and the fundamental interests of humankind, and abide by national and regional ethical norms. Adhere to the priority of public interests, promote human machine harmony, improve people’s livelihood, enhance the sense of happiness, promote the sustainable development of economy, society and ecology, and jointly build a human community with a shared future. (2) Promoting fairness and justice. Adhere to shared benefits and inclusivity, effectively protect the legitimate rights and interests of all relevant stakeholders, promote fair sharing of the benefits of AI in the whole society, and promote social fairness and justice, and equal opportunities. When providing AI products and services, we should fully respect and help vulnerable groups and underrepresented groups, and provide corresponding alternatives as needed. (3) Protecting privacy and security. Fully respect the rights of personal information, to know, and to consent, etc., handle personal information, protect personal privacy and data security in accordance with the principles of lawfulness, justifiability, necessity, and integrity, do no harm to the legitimate rights of personal data, must not illegally collect and use personal information by stealing, tampering, or leaking, etc., and must not infringe on the rights of personal privacy. (4) Ensuring controllability and trustworthiness. Ensure that humans have the full power for decision making, the rights to choose whether to accept the services provided by AI, the rights to withdraw from the interaction with AI at any time, and the rights to suspend the operation of AI systems at any time, and ensure that AI is always under meaningful human control. (5) Strengthening accountability. Adhere that human beings are the ultimate liable subjects. Clarify the responsibilities of all relevant stakeholders, comprehensively enhance the awareness of responsibility, introspect and self discipline in the entire life cycle of AI. Establish an accountability mechanism in AI related activities, and do not evade liability reviews and do not escape from responsibilities. (6) Improving ethical literacy. Actively learn and popularize knowledge related to AI ethics, objectively understand ethical issues, and do not underestimate or exaggerate ethical risks. Actively carry out or participate in the discussions on the ethical issues of AI, deeply promote the practice of AI ethics and governance, and improve the ability to respond to related issues.   4. The ethical norms that should be followed in specific activities related to AI include the norms of management, the norms of research and development, the norms of supply, and the norms of use.

Published by National Governance Committee for the New Generation Artificial Intelligence, China in Ethical Norms for the New Generation Artificial Intelligence, Sep 25, 2021

· 1. THE MAIN PRIORITY OF THE DEVELOPMENT OF AI TECHNOLOGIES IS PROTECTING THE INTERESTS AND RIGHTS OF HUMAN BEINGS COLLECTIVELY AND AS INDIVIDUALS

1.1. Human centered and humanistic approach. In the development of AI technologies, the rights and freedoms of the individual should be given the greatest value. AI technologies developed by AI Actors should promote or not hinder the realization of humans’ capabilities to achieve harmony in social, economic and spiritual spheres, as well as in the highest self fulfillment of human beings. They should take into account key values such as the preservation and development of human cognitive abilities and creative potential; the preservation of moral, spiritual and cultural values; the promotion of cultural and linguistic diversity and identity; and the preservation of traditions and the foundations of nations, peoples and ethnic and social groups. A human centered and humanistic approach is the basic ethical principle and central criterion for assessing the ethical behavior of AI Actors, which are listed in the section 2 of this Code. 1.2. Respect for human autonomy and freedom of will. AI Actors should take all necessary measures to preserve the autonomy and free will of a human‘s decision making ability, the right to choose, and, in general, the intellectual abilities of a human as an intrinsic value and a system forming factor of modern civilization. AI Actors should, during AIS creation, assess the possible negative consequences for the development of human cognitive abilities and prevent the development of AIS that purposefully cause such consequences. 1.3. Compliance with the law. AI Actors must know and comply with the provisions of the legislation of the Russian Federation in all areas of their activities and at all stages of the creation, development and use of AI technologies, including in matters of the legal responsibility of AI Actors. 1.4. Non discrimination. To ensure fairness and non discrimination, AI Actors should take measures to verify that the algorithms, datasets and processing methods for machine learning that are used to group and or classify data concerning individuals or groups do not intentionally discriminate. AI Actors are encouraged to create and apply methods and software solutions that identify and prevent discrimination based on race, nationality, gender, political views, religious beliefs, age, social and economic status, or information about private life. (At the same time, cannot be considered as discrimination rules, which are explicitly declared by an AI Actor for functioning or the application of AIS for the different groups of users, with such factors taken into account for segmentation) 1.5. Assessment of risks and humanitarian impact. AI Actors are encouraged to assess the potential risks of using an AIS, including the social consequences for individuals, society and the state, as well as the humanitarian impact of the AIS on human rights and freedoms at different stages, including during the formation and use of datasets. AI Actors should also carry out long term monitoring of the manifestations of such risks and take into account the complexity of the behavior of AIS during risk assessment, including the relationship and the interdependence of processes in the AIS’s life cycle. For critical applications of the AIS, in special cases, it is encouraged that a risk assessment be conducted through the involvement of a neutral third party or authorized official body when to do so would not harm the performance and information security of the AIS and would ensure the protection of the intellectual property and trade secrets of the developer.

Published by AI Alliance Russia in Artificial Intelligence Code of Ethics, Oct 26, 2021