How governments can seize the GenAI opportunity

Generative AI (GenAI) took the world by storm, mainly with the release of ChatGPT in late 2022. This next generation of artificial intelligence (AI) can generate content — such as text, images and videos — across domains without much data or human guidance, based on a given prompt or context.

GenAI is a powerful tool for organisations. It is not only a technological innovation but also a socioeconomic one that impacts organisations and citizens. But like any new technology, the adoption of GenAI comes with opportunities and risks. How can governments take the lead to embed GenAI into their workforce and transform the business ecosystem by leveraging this technology? 

Support business readiness

Beyond enhancing efficiency and productivity, GenAI is also a catalyst for transformation. By leveraging its capabilities, such as learning, reasoning and creativity, organisations can create new value propositions, enhance customer experiences and generate novel insights. For example, GenAI can enable organisations to personalise products and services at scale, automate complex and creative tasks, and discover new patterns and opportunities in data. 

To achieve these outcomes, they will need to operate in an environment where key resources and opportunities are available, especially talent, infrastructure and ecosystems.

These three elements are interrelated and mutually reinforce and shape one another. For instance, talent can drive innovation and demand for infrastructure and ecosystems, whilst infrastructure and ecosystems can attract and nurture talent. Governments that invest in these elements will be in a great position to attract organisations to operate in their markets, further strengthening the flywheel effect.

Enable the workforce

GenAI has the potential to transform the way people work, learn and collaborate. According to a 2023 report by the National Bureau of Economic Research, GenAI can help upskill employees and improve their performance, especially for new or lower-performing workers. The report also estimates that GenAI could boost productivity by 14% in a contact centre. This is only one of the many examples where GenAI can help increase productivity through a personalised AI assistant.

By providing personalised and adaptive learning experiences, feedback, and guidance, GenAI can help employees acquire new skills and knowledge, improve their problem-solving and decision-making abilities, and enhance creativity and innovation. GenAI can also augment human capabilities by automating routine tasks, optimising workflows, and generating insights and recommendations. Whilst some of these are also true with traditional AI, the broader capabilities of GenAI and its easy interaction with users through natural language will make GenAI assistants much more prevalent. By augmenting existing capabilities, GenAI can enable employees to focus on higher-value and more meaningful work as well as foster stronger collaboration and communication among teams and stakeholders.

GenAI will reshape the nature and scope of work for many occupations and industries. Employees need to be flexible, resilient and willing to learn new skills and take on new challenges as the technology empowers them. Whilst the EY 2023 Work Reimagined Survey revealed that 94% of employers in Southeast Asia are already using or planning to use GenAI within the next year, only 25% of them intend to provide GenAI-related skills training. 

To effectively tap into GenAI, employees need to develop new skills that are essential for working with and alongside intelligent machines. One of these is the ability to confidently leverage AI tools to ask good questions, also referred to as “prompts”. Employees will need critical thinking to evaluate the results and implications of AI outputs. 

Given the magnitude and pace of change that GenAI will bring, governments have a key role to play in fostering a resilient and inclusive workforce that can thrive in the era of GenAI.  They can do this by making sure that training and re-skilling opportunities are available to workers at all levels and stages of their careers. Governments can also partner with educational institutions, employers and other stakeholders to design and deliver relevant, effective curricula and programs that equip learners with the skills and mindsets needed for GenAI. In addition, governments need to promote a culture of lifelong learning and continuous improvement among workers as well as provide incentives and support for them to pursue their learning goals.

Invest in innovation

R&D is essential to advancing the theoretical and practical foundations of GenAI. However, R&D activities are often costly, risky and uncertain, requiring a long-term vision, collaboration and experimentation. This is even more so now with GenAI models — also known as foundation models — which require a significant investment in computing power.  A crucial part of building the models, model training can only be done with graphics processing units (GPUs), which are specialised chips that perform numerous parallel calculations. We are seeing the emergence of private AI labs that often have access to GPUs that are now in high demand and very costly. Some government initiatives, such as the National Supercomputing Centre in Singapore, have started to invest in this area. 

The government plays a crucial role in mobilising and allocating public and private resources for GenAI as it can influence the direction, pace and impact of innovation. It can provide funding, incentives and guidance for R&D, either directly through public institutions and programs or indirectly through grants and partnerships with private actors. Investing in GenAI infrastructure and regulating it can make the technology available, reliable and affordable for different stakeholders. The government can also foster and oversee the data ecosystem for GenAI, promoting its creation, sharing and use, while respecting stakeholders’ ownership, privacy and ethics.

Mitigate risks 

As governments embrace the opportunities and benefits afforded by GenAI, they should also be aware of some potential ethical, social and legal challenges. For instance, how can they be confident of the integrity and reliability of GenAI systems that can generate novel and unexpected outputs? How can they avoid or mitigate the bias that may arise from models trained on biased data? How can they protect the privacy and security of individuals and organisations interacting with GenAI systems or impacted by them? Also, how can governments foster trust and accountability of developers and users of GenAI systems?

To help address these issues, governments can establish guidelines and frameworks that foster responsible development and deployment of AI, such as Singapore’s Model AI Governance Framework by the Infocomm Media Development Authority. In addition, the AI Verify Foundation has brought together many players — including Big Tech companies, startups and system implementers — to help define mitigation approaches to new GenAI risks.  

Another example is KORIKA, an Indonesian government organisation focused on AI research and innovation, which announced a partnership with OpenAI to develop AI models aligned with Indonesian cultural and societal values.2 [2]

These initiatives provide the principles, best practices and assessment tools for organisations to implement trustworthy AI solutions that are human-centric, transparent, responsible and aligned to market needs. By adopting guidelines for such initiatives, organisations can demonstrate their commitment to ethical AI, build reputation and trust among customers and other stakeholders as well as comply with relevant regulations and standards. Moreover, guidelines can help them avoid potential AI risks, such as bias, discrimination, manipulation or misuse that could undermine organisational objectives and values. 

These guidelines can also enable organisations to innovate and experiment with AI safely and responsibly whilst learning from feedback and improvement opportunities. The Association of Southeast Asian Nations has drafted a guide for AI ethics and governance, which is expected to be finalised at the end of January 2024.3[3]

What governments can do next 

Beyond anticipating and preparing the GenAI ecosystem for organisations that operate within their borders, governments need to consider the potential of GenAI to enhance public service delivery and policymaking. Using GenAI to solve complex problems can lead to the creation of new and better solutions as well as generate value and insights for the public.

Governments with firsthand experience in developing and using GenAI can better support citizens and organisations operating in their jurisdictions. For example, Singapore released a writing assistant prototype for civil servants and is expanding the use of GenAI in citizen support chatbots. Other governments in Southeast Asia are considering the establishment of sandboxes to test GenAI use cases for ministries and agencies.

GenAI is an emerging and evolving phenomenon that will have significant implications for the public sector and society. Its initiatives will help develop an ecosystem that includes stakeholders such as academia, startups, accelerators and investors. Government initiatives can also drive greater availability of talent and other resources, allowing businesses to better leverage GenAI to transform the organisation.

Ritin Mathur is a partner, Consulting at Ernst & Young Advisory, and Marie-Claude Ferland is an associate partner, Consulting, AI Strategy, at Ernst & Young Advisory.