Safe and Ethical AI (SEA) Platform Network · Linking Artificial Intelligence Principles (LAIP)

Fairness Obligation

The Fairness Obligation recognizes that all automated systems make decisions that reflect bias and discrimination, but such decisions should not be normatively unfair. There is no simple answer to the question as to what is unfair or impermissible. The evaluation often depends on context. But the Fairness Obligation makes clear that an assessment of objective outcomes alone is not sufficient to evaluate an AI system. Normative consequences must be assessed, including those that preexist or may be amplified by an AI system.

Principle: Universal Guidelines for AI, Oct, 2018

Published by Center for AI and Digital Policy

Related Principles

Proportionality and harmlessness.

It should be recognised that AI technologies do not necessarily, in and of themselves, guarantee the prosperity of humans or the environment and ecosystems. In the event that any harm to humans may occur, risk assessment procedures should be applied and measures taken to prevent such harm from occurring. In other words, for a human person to be legally responsible for the decisions he or she makes to carry out one or more actions, there must be discernment (full human mental faculties), intention (human drive or desire) and freedom (to act in a calculated and premeditated manner). Therefore, to avoid falling into anthropomorphisms that could hinder eventual regulations and or wrong attributions, it is important to establish the conception of artificial intelligences as artifices, that is, as technology, a thing, an artificial means to achieve human objectives but which should not be confused with a human person. That is, the algorithm can execute, but the decision must necessarily fall on the person and therefore, so must the responsibility. Consequently, it emerges that an algorithm does not possess self determination and or agency to make decisions freely (although many times in colloquial language the concept of "decision" is used to describe a classification executed by an algorithm after training), and therefore it cannot be held responsible for the actions that are executed through said algorithm in question.

Published by OFFICE OF THE CHIEF OF MINISTERS UNDERSECRETARY OF INFORMATION TECHNOLOGIES in Recommendations for reliable artificial intelligence, Jnue 2, 2023

Human Determination

The Right to a Human Determination reaffirms that individuals and not machines are responsible for automated decision making. In many instances, such as the operation of an autonomous vehicle, it would not be possible or practical to insert a human decision prior to an automated decision. But the aim remains to ensure accountability. Thus where an automated system fails, this principle should be understood as a requirement that a human assessment of the outcome be made.

Published by Center for AI and Digital Policy in Universal Guidelines for AI, Oct, 2018

· Transparency

As AI increasingly changes the nature of work, workers, customers and vendors need to have information about how AI systems operate so that they can understand how decisions are made. Their involvement will help to identify potential bias, errors and unintended outcomes. Transparency is not necessarily nor only a question of open source code. While in some circumstances open source code will be helpful, what is more important are clear, complete and testable explanations of what the system is doing and why. Intellectual property, and sometimes even cyber security, is rewarded by a lack of transparency. Innovation generally, including in algorithms, is a value that should be encouraged. How, then, are these competing values to be balanced? One possibility is to require algorithmic verifiability rather than full algorithmic disclosure. Algorithmic verifiability would require companies to disclose not the actual code driving the algorithm but information allowing the effect of their algorithms to be independently assessed. In the absence of transparency regarding their algorithms’ purpose and actual effect, it is impossible to ensure that competition, labour, workplace safety, privacy and liability laws are being upheld. When accidents occur, the AI and related data will need to be transparent and accountable to an accident investigator, so that the process that led to the accident can be understood.

Published by Centre for International Governance Innovation (CIGI), Canada in Toward a G20 Framework for Artificial Intelligence in the Workplace, Jul 19, 2018

2. Right to Human Determination.

All individuals have the right to a final determination made by a person. [Explanatory Memorandum] The Right to a Human Determination reaffirms that individuals and not machines are responsible for automated decision making. In many instances, such as the operation of an autonomous vehicle, it would not be possible or practical to insert a human decision prior to an automated decision. But the aim remains to ensure accountability. Thus where an automated system fails, this principle should be understood as a requirement that a human assessment of the outcome be made.

Published by The Public Voice coalition, established by Electronic Privacy Information Center (EPIC) in Universal Guidelines for Artificial Intelligence, Oct 23, 2018

4. Fairness Obligation.

Institutions must ensure that AI systems do not reflect unfair bias or make impermissible discriminatory decisions. [Explanatory Memorandum] The Fairness Obligation recognizes that all automated systems make decisions that reflect bias and discrimination, but such decisions should not be normatively unfair. There is no simple answer to the question as to what is unfair or impermissible. The evaluation often depends on context. But the Fairness Obligation makes clear that an assessment of objective outcomes alone is not sufficient to evaluate an AI system. Normative consequences must be assessed, including those that preexist or may be amplified by an AI system.

Published by The Public Voice coalition, established by Electronic Privacy Information Center (EPIC) in Universal Guidelines for Artificial Intelligence, Oct 23, 2018