Opinion: With AI, Singapore puts the ‘right’ in copyright
20 February 2025

Singapore’s rise as a global hub for commerce and innovation is no coincidence. Decades of thoughtful policymaking have laid the groundwork for its economic growth, establishing Singapore as a hub for technological advancement.
At a time when other countries in the Asia-Pacific region and around the world are calibrating their intellectual property frameworks to better facilitate artificial intelligence (AI) research and development, Singapore is a step ahead. Notably, Singapore’s update to the Copyright Act back in 2021 provided a head start in unlocking AI’s potential while protecting creator’s rights.
Building the future of AI
Developing AI systems is not easy. It requires tremendous computing power, engineering expertise and strategic foresight to produce this technology that can transform businesses, organizations and governments. But above all, AI development depends on access to vast quantities of high-quality data. Without this, AI systems trained on insufficient data are more likely to produce flawed or biased outcomes and inaccurate analysis – think discriminatory hiring algorithms and unreliable medical diagnoses.
When a company sets out to develop an AI system, it breaks data down into small, machine-readable units known as “tokens.” These are put through computational analysis to find mathematical probabilities and correlations across vast datasets. For copyrighted works, this means that when they are included in these datasets, they are not being used for their creative or expressive content but as raw material for analysis.
How a copyright reform powers innovation
Singapore has long maintained an inviting environment for the IT industry, as evidenced by initiatives like the first substantial guidelines for responsible AI. Singapore’s 2021 update to the Copyright Act built on this legacy by introducing the computational data analysis (CDA) exception to allow for the use of copyrighted materials in machine learning.
Crucially, the CDA exception applies to both commercial and non-commercial uses, but only if the copyrighted material is lawfully accessed. This means, for example, that one should not circumvent paywalls to access the material. (Following public consultations in 2024, the Ministry of Law and Intellectual Property Office of Singapore published a report confirming that rights holders can continue to use technological measures to control access to their material.)
In adopting the CDA exception, Singapore was especially prescient. By addressing a set of AI-related issues before the end of 2022, Singapore positioned its economy for future growth ahead of the surge in public awareness about the technology. By offering legal certainty, the exception also encourages businesses to invest confidently in AI training, research and development.
All of this continues to work together to drive the creation of innovative AI products and solutions for incorporation into existing business systems. For example, in cybersecurity, AI systems can sift through large volumes of information created by users of a company’s IT network to forecast, detect, prevent and respond to threats. AI systems can also distill large amounts of data about security events into concrete actions to help companies secure their products and services.
A model for the world: Singapore’s leadership on AI and copyright
Singapore’s policy leadership in this field plays an important role in positioning the country as an attractive base for AI development. Singapore has long championed initiatives that promote the development of cutting-edge technology while attracting businesses and creating jobs. The CDA exception is a standout example of that.
With the CDA, Singapore has become a reference point for other governments considering similar policy solutions. In the Asia-Pacific region, for example, both Hong Kong and South Korea are actively considering introducing CDA exceptions of their own to accelerate growth in their AI industries.
AI governance is an important issue that highlights broader policy questions: To what extent can existing laws address AI-related issues? What policy changes or new laws are needed to widely spread the benefits of AI while tackling some of its novel challenges?
To address these questions, government and industry leaders should look at how Singapore has handled its copyright law to protect the interests of rights holders while encouraging AI innovation.
- Jared Ragland, Senior Director of Policy for Asia Pacific, Business Software Alliance