Skip to main content
How Web3 and AI Will Transform Finance

Economics & Finance

How Web3 and AI Will Transform Finance

How Web3 and AI Will Transform Finance

Artificial intelligence and crypto are not only reinventing financial products and delivery, but also influencing who gets to participate.

Since the birth of ChatGPT, writing summaries from long reports or crafting personalised emails have become a thing of the past for many people. One year on, the ability of generative AI (GenAI) to extract and synthesise data and sentiments from massive volumes of data has made complex information more accessible than ever before. It is now also possible to tailor one-to-one online customer interactions at scale, from product search to purchase. 

The exponential increase in operational and cost efficiencies made possible by GenAI is observed across domains – including the way financial services are delivered and experienced by consumers. Its ability to personalise market intelligence and the customer journey in a fraction of the time presents a huge growth opportunity, said Rishi Ramchandani, the APAC Web3 Lead at Google. 

This comes at a time when the finance world is being disrupted by yet another formidable force: Web3 technologies, or what is commonly known as blockchain. In the INSEAD Future Forum themed “How Will Web3 and AI Transform Finance?”, practitioners took a deep dive into how these technologies are shaping finance as we know it. 

Heralding a new era

The potential of GenAI to drive cost and operational efficiencies, as well as mass customise interactions at scale is unprecedented, concurred experts on the “Generative AI: A Game Changer for the Future of Finance” panel discussion. In an increasingly uncertain commercial and regulatory landscape, it can help developers and analysts make sense of regulatory changes and identify or even automate the required changes. For instance, GenAI can distil the essence of the Basel Framework that is relevant to the business and identify related parts of a product document and code repository where change is required.

The fact that GenAI processes text, as well as image, voice and video not only makes a variety of use cases possible, but also does a lot to increase inclusiveness, said Alex Honchar, Director of AI Engineering and Partner at Neurons Lab. Indeed, beyond the tangible changes, there are philosophical ones – including financial inclusiveness, noted Stefano Bury, INSEAD alumnus (MBA’13D) and COO of LongHash Ventures

With blockchain, everyone and anyone can now access financial services from anywhere. In the past, huge sums of money were often locked up in financial markets due to middlemen and other forms of operational friction. But the decentralised, community-based and -owned nature of blockchain technology makes it possible to bypass traditional, institutionally owned systems – together with their associated fees, rules on minimum transaction volume and operational lag. 

In the panel discussion “How Will Web3 Shape Finance 2.0”, it is clear that the reduced cost, speed (think: instant settlement) and improved accessibility are a big win for financial inclusivity.

Moreover, digital assets and the underlying blockchain technology are giving rise to new asset classes. Following in the footsteps of cryptocurrency, real-world assets such as mortgages are being tokenised. Digital assets such as Bitcoin could also serve as economic instruments to hedge inflation (like gold) or raise venture capital, said Samar Sen (INSEAD MBA’08D), Senior VP and Head of APAC at Talos. However, he cautioned that the maturity levels of different use cases vary, and it is too early to tell which will be successful. 

Harald Eltvedt, INSEAD alumnus (GEMBA’20J) and venture builder at SC Ventures asked: “We can tokenise everything, but does everything need to be tokenised?” His answer was “probably not”. 

Are we there yet? 

As these relatively nascent technologies continue to unfold, like many panellists, Hassan Ahmed, Country Director for Singapore at Coinbase, pointed to scalability being a challenge, primarily due to the lack of interoperability and market stability.

Interoperability requires market players, as well as the relevant stakeholders, to operate using common rules, standards and systems. Literally and figuratively, it calls for a common language.

Second, market stability is a necessary ingredient for trust and increased adoption. To ensure a more stable global market, regulatory collaboration and joint oversight can go a long way. We are now seeing cross-border cooperation between regulatory communities, such as between Europe and Singapore, Asia and Europe, as well as US to Singapore and Africa, said Anton Ruddenklau, Head of Financial Services at KPMG Singapore. For consumer trust and protection, liquidity reserves can be implemented as a safety net. There will be shocks in the systems and we need to build trust and liquidity in these systems, he said.

With little standardisation, market stability and regulatory certainty, progress at a global scale will look nothing more than patchwork. In the meantime, pension funds and big players will simply wait at the sidelines.

Four steps to business transformation 

Ramchandani highlighted four steps for companies to better leverage the potential of GenAI. In fact, these steps are applicable to Web3 or other emerging technologies.

First, identify possible use cases and prioritise them. Leaders need to evaluate business needs alongside generative and traditional AI capabilities. Taking a realistic perspective of the time to value for each use case will help companies determine the feasibility of each one.

Second, once they have ascertained their priorities, firms should take a “data first” strategy. Lay a solid AI foundation by bringing together disjointed data sets and strong governance. To ensure that AI models are robust and serve the purpose they are built for, organisations must audit both the AI models and training data, said Laurence Liew, Director of AI Innovation at AI Singapore. A case in point is how OpenAI spends most of its resources on humans checking the outputs of their models. 

Third, integration and implementation of AI solutions can only be successful if the organisation is ready. Thus, leaders need to look beyond the cost and benefits of GenAI solutions and consider the organisation’s AI readiness. Like most panellists, Zaid Hamzah, Managing Partner of, agrees on the need for learning and reskilling, and along with that, the need to strike a balance between dependence and professional judgement. He said that investing in people not only helps companies reap the return of investment into GenAI, but also ensures that everyone can benefit from GenAI. 

Fourth, recognise that not all AI is built equal. Test proofs of concept to identify what works for the business before scaling them. A deeper understanding of how the technology works and its capabilities can help decision makers move from asking “Where can we use it?” to “How good is it right now?”, said LingYi Chang, Founder and AI Consultant at 01 Solar. 

Integrating GenAI will continue to raise these questions: How do we control our data? How to be both accurate and explainable? How to integrate our existing data and applications? How to control costs? How to deal with fraud and security? While some of these issues can be tackled at the company level, some require an ecosystem effort, or even regulatory intervention.

Where do we go from here?

From the perspective of governance and regulation, the question is: How to create guardrails and ensure responsible use? In the case of GenAI, audit is most critical at the foundational model stage, where the model runs in a controlled environment, remarked Vihang Patel, INSEAD alumnus (CGM’21) and Founder and CEO of Splore. It is much more challenging to regulate or provide guardrails “after the fact”, he said. Therefore, the perfect time is sooner, rather than later. 

In addition, a different approach is required in the regulation of GenAI. Current regulations tend to focus on the inputs and outputs of AI models, whereby regulations are based on the size of model and responsible use of outputs. But we need to go beyond the size of model to look into human responsibility in the use of GenAI, such as the issue of bias

On the Web3 front, in spite of the recent collapse of FTX amid other crypto fallouts, Sen noted that safety has in fact improved in the last five years. He highlighted the “good” that came out of recent disasters: winding up of scamming companies, increased consolidation in the sector – and more importantly, a better understanding of the need to separate custodian and trading. The benefits of blockchain allowing everyone and anyone to access financial services come with an urgent need for consumers to better educate themselves, he cautioned.

At what point then would we have “arrived”? In an ideal, abstract world, transactions would take place without the need for external intervention, Sen projected. In any case, there’s no turning back. “We are on this ship together and there is nowhere to escape,” said Chang.

Edited by:

Geraldine Ee

About the author(s)

Related Tags

Artificial intelligence

About the series

AI: Disruption and Adaptation
Delve deeper into how artificial intelligence is disrupting and enhancing sectors – including business consulting, education and the media – and learn more about the associated regulatory and ethical issues.
View Comments
No comments yet.
Leave a Comment
Please log in or sign up to comment.