AI in general 1, 2018

Statement on Artificial Intelligence, Robotics and ‘Autonomous’ Systems

European Union

Actors

European Group on Ethics in Science and New Technologies

Tags

Accountability, Bias, Sustainability, Responsibility, Autonomy, Labour market effects, Privacy,

Resources


The manuscript is issued by the European Group on Ethics in Science and New Technologies, which is an independent advisory body of the President of the European Commission. The paper consistently refers to itself as being a “statement” and has the aim of laying down the basic moral principles for future international law on the use and governance of artificial intelligence, robotics, and autonomous systems. It also believes that the EU must play a defining and leading role in the crafting of this more comprehensive international law.
The issuing of the statement was motivated by the ever-growing “opaque” mode of operation of the various autonomous technologies, and by the fear of a fragmented or unharmonized way of AI regulation (“regulatory patchwork”) within the European Union. This latter was the more important consideration behind the globally focused statement, since the basic principles and democratic prerequisites that the manuscript sketches are “based on the fundamental values laid down in the EU Treaties and in the EU Charter of Fundamental Rights.”
These ethical principles and democratic prerequisites are: 1) human dignity; 2) autonomy; 3) responsibility; 4) justice, equality, and solidarity; 5) democracy; 6) rule of law and accountability; 7) security, safety, bodily, and mental integrity; 8) data protection and privacy; 9) sustainability.
1) The principle of human dignity should limit the use of AI in respect to making classifications and determinations about human beings based on algorithms and autonomous technologies. Human beings should be informed about the use of autonomous technologies when they could be affected by their application. Persons should also know, whether they are interacting with an actual human being or with a machine.
2) The principle of autonomy prescribes that human beings should have control over and knowledge about autonomous systems. Human beings should be able to set their own norms and standards and to live a life based on the principles they established for themselves. Persons should be able to make their own decisions and take actions alone. However, if these actions and decisions were delegated to autonomous technologies, humans should be in absolute control over their “when” and “how.” This entails that autonomous systems should be transparent and predictable.
3) The principle of responsibility should be the guiding principle for AI research and application. This means that future development should align with the global social and environmental preferences of the nations of the world, which should be defined through a democratic process. Fundamental human values and rights should never be endangered, and the application of AI and robotics should never “pose unacceptable risk of harm” and should not “compromise human freedom and autonomy,” especially by limiting options for, and awareness of citizens.
4) The principles of justice, equality, and solidarity entail that AI should contribute to global justice with benefits accessible for all. To achieve this, first, the creation of biased data sets for AI systems should be prevented. Second, the global accessibility to core AI technologies should be ensured in order to provide fair distribution and benefit sharing coming with the hardships of the economic transformations that automation, digitalization, and AI bring.
5) The future regulation of AI development and application should be the outcome of deliberative democratic debates because only this can guarantee an “inclusive, informed, and farsighted manner.” The pluralism and diversity of democratic societies “must not be jeopardized, subverted, or equalized by new technologies,” therefore, the freedom of expression and the right to receive information should always be guaranteed.
6) To implement in practice the concept of and to protect the different human rights, the rule of law and accountability must always prevail. Within the developing new field of AI regulation, governments must invest in the creation of “robust solutions” that ensure a “fair and clear” distribution of responsibilities.
7) When it comes to the principles of security, safety, bodily and mental integrity in relation to autonomous systems, the following three forms are given extra care by the statement: first, the external safety; second, reliability and internal robustness; and three, emotional safety with respect to human-machine interaction. These areas should be taken into account by AI developers before release. Special attention should also be paid to people in a vulnerable position and to the weaponization of AI.
8) Physical AI robots as part of the Internet of Things and AI softbots that operate via the World Wide Web must comply with data protection regulations and not spread and collect data without informed consent in order to protect personal information and the right to privacy. The manuscript enlists three rights that belong to the right of private life and should be respected by the operation of autonomous systems: first, the right to be free from technologies that influence personal development and opinions; second, the right to establish and develop relationships with other human beings; and third, the right to be free from surveillance. It also opens the debate, whether the right to private life should be enriched by the following two new rights: the right to meaningful human contact and the right not to be profiled, measured, analyzed, coached or nudged.
9) Priority must also be given to policies of environmental protection and sustainability when designing future AI regulation, since humans have a responsibility to ensure basic the preconditions for life on the planet.

AI Governance
Database

This database is an information resource about global governance activities related to artificial intelligence. It is designed as a tool to help researchers, innovators, and policymakers better understand how AI governance is developing around the world.

We're collecting more examples as we find them. If you'd like to add yours to this map, please suggest an entry.

Suggest an entry
  • Filter by:
  • Current filters:
256 entries are listed below
  • Sort by: