Standards Development

Scope

This document specifies the requirements and provides guidance for establishing, implementing, maintaining and continually improving an artificial intelligence management system within the context of an organization.

This document is intended for use by an organization providing or using products and/or services which use artificial intelligence. This document helps the organization to develop or use artificial intelligence responsibly in pursuing its objectives, and to meet applicable regulatory requirements, obligations related to interested parties and expectations from them.

This document is applicable to any organization, regardless of size, type and nature, which provide or use products and services using artificial intelligence.

Purpose

The purpose of the proposed standard is to enable an organization to show that it has implemented and continually iterates on the improvement of processes unique to the development or use of an AI system, for example processes identifying and treating bias of learning data, or more generally fairness, inclusiveness, safety, security, privacy, accountability, explainability and transparency, etc.

With an increasing rate, organizations world-wide are developing and using products and services that are based on artificial intelligence (AI) for purposes of analysis, deriving information, decision making or automation. AI is about to become an ubiquitous technology used in many sectors and in all types of systems, without sometimes the users being aware of it. There are opportunities and threats associated with the use of AI. Various concerns related to applications of AI and its impact on entire sectors and the society. The expressed concerns include, amongst others, the following:

- fairness: how to ensure that people impacted by the use of AI are impacted in a fair manner with regard to their expectation; how to avoid discrimination (for example, AI based advertising system not showing some job advertising to women)

- automated decision-making having impact on people: how to ensure that the decision making comply with expectations from stakeholders, including regulators.

- transparency and explainability of AI system: how stakeholders can have the expected information on the way AI systems operate

- human oversight: how humans can intervene in the AI system

- impacts on safety.

The sectors which are impacted include, among others, the following:

- health

- defence

- transport

- finance

- employment

- energy.

Due to the wide impact, governments and regulators worldwide are discussing the use of AI. They are defining AI-related guidelines and there are already intentions to regulate AI.

The MSS for AI can help organizations to assess the AI-related opportunities and threats and address the raised concerns, thereby supporting them in achieving their objectives. Showing conformity to the AI MSS can improve the acceptance of AI solutions by the society and help to meet applicable  regulatory requirements and governmental guidelines.

The MSS for AI is intended to be used independently or complementarily to existing management systems standards.

Consumers in turn can rely on certification to increase trust in the products and services that use AI provided by large industry or SMEs.

Governments and policy makers can use the proposed standard for possible 3rd party certification to provide a basis for regulation of a large number of industry sectors, the proposed risk-based standard being more flexible and applicable than prescriptive sector specific requirements.