The European Commission (EC) released its Communication COM(2018) 237 ‘Artificial Intelligence for Europe’ on 25 April 2018, setting out a European initiative on Artificial Intelligence (AI), which is part of its “Delivering on the Digital Single Market – 3rd Data package”. This awaited initiative is essential for the future European economy and the leadership of our industry in a competitive global market. CEN and CENELEC welcome the recognition of standardization in contributing to answer the challenges this key technology brings, particularly in terms of safety and liability and ethical considerations.
Artificial Intelligence (AI) is progressing continuously and is affecting widely the industry e.g. automation, data management, integration of smart technology. Society is also impacted as AI changes the way businesses operate, streamline production and transforms workers' skillsets. AI applies to a variety of sectors where standardization is of high relevance: smart manufacturing, robots, autonomous cars, virtual reality, healthcare, interactive speech interfaces, visual recognition, data analysis/manipulation, home appliances, cybersecurity or spatial programming. Challenges identified in the EC Communication refer to the deployment, interoperability, scalability, safety and liability of AI, thus creating a need for standardization.
How European standardization benefits the deployment of AI
European standardization is an essential element to avoid fragmentation in the Digital Single Market and ensure a people-centric approach to AI by making sure that it is benefitting society. CEN and CENELEC standards create an environment of trust and accountability around the development and use of AI technology, addressing issues such as reliability, safety, privacy and security. Moreover, CEN and CENELEC have strong ties with their international standardization bodies, the International Organization for Standardization (ISO) and the International Electrotechnical Standardization (IEC), where work already kicked-off on AI in ISO/IEC JTC 1 SC 42 (learn more below).
European standardization within the proposed EU initiative on AI
The AI initiative of the European Commission underlines the “solid basis of standards” already existing in the area of AI-enabled devices related among others to the Machinery Directive, the Radio Equipment Directive, the EU safety legislation for medical devices or toys, as well as the Product Liability Directive. The Communication recognises that “The further development and promotion of such safety standards and support in EU and international standardisation organisations will help enable European businesses to benefit from a competitive advantage, and increase consumer trust”.
European Standardization Bodies are ready to take on board standardization work in support of AI technologies, as was already underlined by the CEN and CENELEC Digital Champion Dr Bernhard Thies: "We are living in a very exciting time, a time of evolution for industry where intelligent manufacturing has to be supported by advanced standards."
Expected AI standardization work at European level
CEN and CENELEC support European legislation in developing Harmonized Standards. The European Commission already foresees impacts in the machinery sector with changes or clearer guidance brought to a series of EU Directives such as the:
This means that some of the 50 CEN Technical Bodies dealing with machinery safety will be impacted, as well as some CENELEC Technical Committees in the field of smart appliances such as:
- CLC/TC 61 ‘Safety of household and similar electrical appliances’
- CLC/TC 59X ‘Performance of household and similar electrical appliances’
- CLC/TC 205 ‘Home and Building Electronic Systems (HBES)’
- CLC/TC 57 ‘Power systems management and associated information exchange’
These CENELEC TCs represent 582 published standards, among which 67% are directly supporting European legislation through Harmonised Standards under the above-mentioned Directives and Ecodesign Regulations.
Expected AI standardization work at international level
At international level, the Joint Technical Committee 1 between ISO and IEC (JTC 1) and its subcommittee (SC) 42 (JTC 1/SC 42) have taken on board standardization work on artificial intelligence under the lead of the American National Standards Institute (ANSI) and US secretariat.
JTC 1 SC 42 counts 18 Members, of which more than half are European and Members of CEN and CENELEC: Austria (ASI), Denmark (DS), Finland (SFS), France (AFNOR), Germany (DIN), Ireland (NSAI), Italy (UNI), Portugal (IPQ), Switzerland (SNV), United Kingdom (BSI).
This newly created Committee will provide guidance to other IEC, and ISO committees that are developing AI applications, looking at the full AI IT ecosystem. Work started on:
- Artificial Intelligence Concepts and Terminology
- Framework for Artificial Intelligence (AI) Systems Using Machine Learning (ML)
The way forward
To help us identify the challenges posed by AI to industries and define a roadmap for AI standardization, CEN and CENELEC are organizing a dedicated ‘stakeholders workshop’ on “Trustworthy AI – building a framework with standardization” on 18 September 2018
CEN and CENELEC welcome this timely initiative form the European Commission, and are looking forward to having an active role in the deployment of AI technologies in order to maintain and ensure European industries’ leadership.
Read the press release from the European Commission on their Communication COM(2018)237 ‘Artificial Intelligence for Europe’