Artificial Intelligence

AI is crucial for the emergence of next-generation products and services contributing to the dynamism, inclusivity, and competitiveness of ASEAN’s digital economies. To this end, AI standards help ensure AI grows and matures in a fair, secure, and interoperable manner.

What is AI?

AI is a general term for computing systems that emulate human cognitive functions, such as identifying patterns to solve problems.
Comprising machine learning, deep learning, big-data analytics, automation, and some types of robotics, AI improves productivity, streamlines processes, and optimises resources. It can also address complex, longstanding socio-economic issues such as hunger, poverty, inequality, discrimination, and climate change.

Key Trends in AI

The way AI is defined and approached is largely fragmented across the ASEAN region, with no single, unifying regional governance framework to facilitate collaboration within and across member states. A harmonised approach to AI is vital for ASEAN to effectively become a global and regional AI powerhouse.

Different facets of AI were used in numerous manners to mitigate the negative effects of the COVID-19 crisis (contact-tracing efforts, virus detection protocols, supply chain rationalisation, etc.). Now that the tourism-led economies of ASEAN are looking at the post-pandemic horizon, it is time to operationalise AI in a more durable manner.

Projected contribution of AI to GDP in ASEAN Economies

Key AI Standards

  • The ISO/IEC 27000 family of standards is widely regarded as a baseline for data protection principles (ISO/IEC 27001 on Information Security Management and ISO/IEC 27002 on codes of practice for information security controls).
  • ISO/IEC 27701 provides specific guidelines for the establishment, implementation, maintenance, and continual improvement of privacy.
  • ISO/IEC 27010 specifically governs the sharing of information across organisations.
  • ISO/IEC 27000 also refers specifically to cloud processes with ISO/IEC 27017.
  • IEEE 2301 addresses cloud interoperability and portability, which can be adopted to prevent uncoordinated implementations of cloud infrastructure or be taken as reference for more specific approaches at the regional level.
  • ISO/IEC SD 5259 on data quality for analytics and machine learning is currently under development.
  • IEEE P2247.1 governs the classification of AI-driven adaptive instructional systems.
  • IEEE P2660.1 covers practices for the integration of low-level automation functions and other software agents in industrial control platforms.
  • ISO/IEC JTC 1/SC 42 working group oversees the development standards around AI ethics such as bias and transparency.

Major initiatives include: ISO/IEC DIS 24668 on Process Management Frameworks for Big Data Analytics, ISO/IEC DIS 23053 on a Framework for AI Systems using Machine Learning, and ISO/IEC NP 38507 on Governance Implications of the Use of Artificial Intelligence by Organisations.

  • IEEE 7000-2021 addresses ethical concerns during system design, among others.
  • ISO/IEC TR 24028:2020 on the trustworthiness of AI, covers issues like transparency, while ISO/IEC TR 24027 addresses bias in AI decision-making.
  • In terms of inclusion and accessibility, ISO/IEC GUIDE 71:2014 addresses accessibility in standards, ISO/IEC 30071-1:2019 covers the accessibility of user interfaces, and ISO 21801-1:2020 tackles cognitive accessibility.

Opportunities and challenges

  • Economic impact – AI enables automation and intelligent insights, leading to increased productivity gains, especially for SMEs.
  • Government efficiency – AI can free up government staff and help overcome budget, resource, and manpower constraints.
  • Social resilience – AI is a powerful tool for multiple sectors, with impact especially important in remote or rural areas.
  • Universal Accessibility of AI – AI can help fight discrimination and contribute to the inclusion of persons with disabilities and marginalised groups.
  • Cross-Border Data Flows – Excessive constraints on data flows can impede AI innovation to spread.
  • Privacy and Data Protection – The personal and organisational data that is vital to AI must be kept safe and secure wherever it is needed.
  • Ethics, Transparency, and Accountability – AI must be developed and implemented in a responsible and accountable manner.
  • Skills, Training, and Connectivity – Workers and businesses need the skills, knowledge, and confidence to both use and benefit from AI.

Comments are closed.