ULG's Language Services Blog

Regulating Big Data: SC 42 and Artificial Intelligence

 

According to the International Data Corporation (IDC), 40% of digital transformation initiatives will use Artificial Intelligence (AI) services by 2019 and 75% of enterprise applications will use AI by 2021. The IDC also estimates that spending for AI services is projected to reach $19.1 billion in 2018 and $52.2 billion by 2021.

With this expected exponential growth of AI, there is an increased need for global organizations to be able to communicate with each other clearly about AI and to use and develop these new technologies responsibly. As a result, in March 2018, the ISO/IEC JTC 1/SC 42 subcommittee (SC) was formed by the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) to provide guidance and standardization for developing Artificial Intelligence applications.

Similar to how other ISO standards provide guiding frameworks for industries as diverse as medical devices and energy management, SC 42 is expected to set forth guidelines for global organizations to follow to ensure best practices when using AI. While SC 42 is still in the initial stages of its research, its impact on industries that rely on technology, including the language services industry, will be extensive and must take into account not only the importance of standardization, but also globalization.

 

What are the goals of ISO/IEC JTC 1/SC 42?

 

According to Wael William Diab, the Chairperson of SC 42, the subcommittee is meant to “serve as the focal point for AI standardization” and become  “a systems integration entity.” The subcommittee is broken into four working groups, each with a specific goal.

The foundational standards working group will create a common terminology for AI concepts to ensure that organizations involved in Artificial Intelligence are using the same language.

The computational approaches and characteristics of Artificial Intelligence systems study group will study different technologies such as Machine Learning algorithms and specialized AI systems to understand their underlying architecture and characteristics. This group will also study current industry practices, including processes and methods for the application of AI systems as a starting point for defining the best practices that will become standardized.

The trustworthiness study group will explore ways to establish trust in AI systems and assess methods to minimize threats and risks to AI systems.

The use cases and applications study group will collect representative use cases with different AI applications, such as social networks and embedded systems, and the different contexts in which they are used, such as healthcare and finance. This group will then describe the use cases using the terminology and concepts defined by the foundational standards working group.

Together, these four groups will produce a series of standards and best practices for the use of AI, as well as technical reports and other necessary deliverables to be used on an international scale.

 

What does this mean for the language services industry?

 

Previous ISO standards have had far-reaching impact for the language services industry. Standards such as ISO 17100 have set forth the processes and industry procedures for ensuring accurate translations, while ISO 13485 and ISO 9001 have outlined requirements for quality management systems when translating materials for medical devices and consumer products, respectively.

SC 42 is expected to have a similar impact on the language services industry. Language Solutions Partners (LSPs) use AI, specifically tools such as translation memory, language glossaries, and machine translation, to reduce the cost of translation and expedite the translation process.

AI is also used in localization practices, such as changing alphabet systems according to the language used and adapting fonts and letter sizes to ensure different languages can fit on the same screens. LSPs will therefore need to comply with any AI standards that SC 42 releases.

As SC 42 delves deeper into its study of the current landscape and begins to form the standards that will govern AI usage, it must also take into account how AI can be used to promote globalization and multilingualism. Current practices and needs of LSPs will be an essential part of this research, and LSPs will therefore have the opportunity to shape AI standards in the future.