The Global Standards Collaboration (GSC) held its 22nd meeting in Montreux, Switzerland on 26-27 March, hosted by IEC and ISO. Over 100 participants representing 12 leading global information and communication technologies (ICT) standards development organizations (SDOs) attended.
The event covered diverse aspects related to smart sustainable cities and artificial intelligence (AI).
“Artificial intelligence offers a lot of promise and will give us a lot of opportunities in terms of how we live, work and play, but it also comes with some challenges, such as trusting and transparently explaining its decisions to the end-user”, said Wael William Diab, who chairs the IEC and ISO joint technical committee on AI (ISO/IEC JTC 1/SC 42), and chaired the GSC session on AI, “SDOs play an important role in addressing these concerns and provide an opportunity for global standards to accelerate adoption of AI in support of the digital transformation.”
Learning algorithms and machine learning are some of the technologies at the heart of artificial intelligence (AI). They are being used increasingly in daily life. For instance, to help healthcare professionals make better decisions for patients and drive higher efficiencies in smart manufacturing, by providing insights for production planning.
The list of applications is numerous and growing, including retail, consumer, financial, digital assistants, connected cars, smart grids, market intelligence and more.
Why the whole AI ecosystem must be considered
AI isn’t a single technology, it’s a collection of technologies with numerous and diverse stakeholders, who are approaching the deployment of AI systems from a business angle, in other words, with a focus on customer needs, segments, services, products and regulatory requirements. While these technologies bring many benefits, they also raise a number of concerns.
“Broadly speaking, these aspects are related to trustworthiness, ethics and societal concerns which must be addressed early on in standards and technology development, so that technology can be deployed widely. This is why SC 42 is taking a very broad approach and looking at the entire AI ecosystem”, said Diab.
The session covered:
- General and overview presentations on AI providing an update on the various work programmes of the SDOs in this area
- Areas related to the implication and concerns of deploying AI, such as societal concerns, ethics, governance and trustworthiness, along with proposed solutions, standards and mitigation
- Application areas, representative use cases and related technologies (big data, analytics), to better understand requirements coming from these domains
- Potential synergies and collaboration opportunities between GSC members
From developed to emerging markets
Participants gave updates on their work, offering diverse perspectives and insights. While some of the challenges differed from regions and countries, the concerns were largely the same.
Yutaka Miyake from the Telecommunication Technology Committee (TTC) talked about the need for transparent and trustworthy AI and machine learning, as Japan implements its “Society 5.0” plan for a super smart society. Miyake highlighted the need to uphold the social principles of AI, which must remain human-centric and accountable, while allowing fair competition for sustainable economic growth and innovation. He also raised challenges around data privacy and security, for example the hackability of facial recognition systems, as well as algorithm bias, which could lead to unfair decisions.
As AI becomes embedded in 5G networks in India, it will help improve agriculture, broaden access to education and healthcare for some 60% of remote populations, and be deployed in smart cities and infrastructure. Presenter Samar Shailendra from the Telecommunications Standards Development Society ( TSDSI ) noted the need to focus on continuity of services and coverage, as well as security”. He also talked about ethics regarding job security and stressed that AI should complement people rather than replacing them.
The next generation of phones will not be smart, they will be intelligent, enabled with AI and combined with next generation communication networks. They will use huge amounts of data and need to access information more accurately, quickly, automatically and intelligently. Thomas Li from the China Communications Standards Association (CCSA) emphasized the importance of getting AI right, noting that standards will need to include AI ethics, security, service application programming interfaces, as well as performance of AI vision, hearing and conversation.
Diab concluded, “it was fantastic to see all these SDOs in one room, where the number of ideas and ongoing projects for AI are very impressive”, adding that “standards provide an open platform for innovation and a venue to address the needs and concerns of diverse stakeholders to help advance some of these strategic initiatives we’ve been talking about these last few days”.
GSC brings together SDOs from around the world, to promote global cooperation for ICT standardization. Since 1990, it has provided a regular exchange of work programmes and other information in different technical areas.
Find out more about GSC-22 and its members.