Nigel Clarke, Deputy Managing Director of the IMF and President of the Per Jacobsson Foundation, said that in the lecture's 62-year history, no co-founder of a company not yet 5 years old had ever taken the stage. He acknowledged that the choice of speaker might look unusual in hindsight, but said the moment the world is living through is not a normal one either.

As AI advances and raises new questions for the global economy, policymakers cannot find the answers alone. Technology developers cannot answer every question by themselves either. Both sides need to be in the same conversation.

That is why the Per Jacobsson Foundation Lecture at the 2026 IMF-World Bank Annual Meetings in Bangkok invited Jack Clark, co-founder of Anthropic, to speak on "Inside the Event Horizon: Why Everyone Should Take AI Progress Seriously".

What Clark presented was a future in which AI is no longer just a tool that helps people work. It becomes a discoverer of new knowledge, an accelerator of scientific progress, and possibly a contributor to building the next generation of AI itself.

He believes the world is reaching a turning point where technological progress will move faster than humans are used to, and that we have less than a decade to prepare.

From Journalist to Anthropic Co-Founder

• Clark opened with a joke that his key qualification is being a human, and that there is only one of him. It took him almost 40 years to build up the experience to stand on this stage.
• He began his career as a journalist covering data centres. His interest in AI grew until he left journalism in 2016 to join OpenAI, and he later co-founded Anthropic.
• Over the years, he has tried to tell the world that one day AI would be capable enough to change national security and the direction of the global economy, and might even help build the next generation of AI.
• Today he believes that moment has arrived, because AI is affecting security and the economy and is taking part in its own development.
• The warning he wants to give is that the change ahead may be far bigger than the world expects.

The Singularity May No Longer Be a Future Event

• Clark brought up the concept of the Singularity to describe a period in which AI can accelerate scientific and technological progress so quickly that humans may not be able to understand it or keep up.
• He cited Vernor Vinge, the science fiction writer who proposed the idea in 1993. Vinge saw that once systems smarter than humans begin to drive progress, those systems could build even smarter ones in ever shorter periods of time.
• At some point, the models humans have used to understand progress may no longer be able to explain the world.
• Clark believes 2026 is the year when credible evidence began to emerge that the Singularity has already begun.
• He expects the progress of AI and science to speed up sharply over the next few years, which could create both immense wealth and serious danger.
• Even so, he chose to begin this lecture with the opportunities and benefits of AI, without denying that the risks are real.

AI Could Change the Rate of Scientific Progress

• Clark sees one of AI's most important benefits arising in science, from discovering new knowledge and developing infrastructure to curing disease.
• Over the past half-century, the number of researchers has grown rapidly, yet researcher productivity has fallen, because new ideas have become harder and more costly to find.
• AI may change this constraint, because it can extend researchers' capabilities and take over some work that once depended on humans.
• The clearest example is software development, where Clark said AI is already more capable and efficient than most of the world's software engineers.
• With AI's help, a single person can now build software that once needed a team of several people and months of work.
• Internal Anthropic data that he presented indicated that the amount of code researchers produce has risen almost 10-fold compared with the company's early days. A slide showed that in the second quarter of 2026, code volume per person was 9.3 times higher than the average before 2025.
• However, progress in coding cuts both ways, because AI that helps build cyber defences can also be used to attack systems.
• He sees the same thing happening in science and robotics, which could change the rate at which the global economy produces new innovation.

First Signs That AI May Be Doing Frontier Science

• Clark said that in 2026 there are the first signs that AI systems may be capable of doing science at the frontier of knowledge.
• In mathematics, AI has been used to propose solutions to long-standing open problems, such as Navier-Stokes, one of the Millennium Prize problems, as well as the Jacobian Conjecture and Erdős problems.
• In many cases, AI did not just help with calculations but contributed key ideas that led to the solution.
• In biology, researchers have begun using AI to design viruses from scratch and to search for compounds with potential to become drugs. Some drugs in whose design AI played a significant role have already moved into testing.
• In robotics, AI can increasingly operate laboratory equipment on its own, allowing experiments in biology, materials science, batteries and fuel cells to move faster than before.
• Clark sees it as significant that these abilities do not come from separate kinds of technology. They can come from the same foundation model.
• He calls systems such as Claude, ChatGPT and Gemini "Virtual Polymaths", capable of doing high-level science across many fields.

AI Exam Scores Leap From Single Digits to Near Perfect

• Clark gave the example of Humanity's Last Exam, a test that gathers hard problems from many fields, from chemistry, computer science and mathematics to classical literature and ecology.
• When it launched in January 2025, the best-scoring AI system managed only 8%.
• In the week of his lecture, Claude Opus 5.5, using tools, scored 67.7%, according to data he presented from Scale AI and CAIS.
• Another example is FrontierMath, a test of hard mathematics problems compiled and checked by professional mathematicians.
• When it launched in November 2024, the best AI system scored under 2%.
• Clark said the top score has since risen to 94%, achieved by GPT-6 Astra, according to Epoch AI.
• He said this pattern is not limited to a few fields. In the week before his lecture, he collected around 100 graphs of AI progress and found a pattern of rapid improvement across many disciplines.
• For him, this is evidence that AI capability is no longer growing slowly and is clearly accelerating.

When AI Found the Scientists' Answers Were Wrong

• Clark described TheoPhysEval, a study published on arXiv in October 2026, which used 50 theoretical physics problems whose answers had never been published.
• The problems were compiled from physicists' research around the world, so AI could not rely on memorising answers that had already appeared on the internet.
• He quoted a Facebook post by Suvrat Raju, one of the study's authors, who wrote that he did not know of any individual who could solve any significant fraction of the final collection of problems without AI assistance.
• Initial results showed GPT-6 Astra scoring in the late 80% range, with Claude Opus 5.5 close behind.
• But when the researchers examined the questions the AI got wrong, they found that the AI, Astra in particular, insisted its own answers were correct and that the experts' answer keys might be mistaken.
• When the experts went back and checked, they found errors hidden in their own calculations in several cases.
• After the answer keys were corrected, Astra's score rose to 98%, or 49 out of 50 questions.
• Clark compared the episode to a scene in films where the apprentice proves to the teacher that they are ready to move forward on their own.
• For him, it shows that AI is not only able to answer questions from what it has learned, but is beginning to show creativity in solving scientific problems that have never had answers.
• If AI can truly create new knowledge, that ability could extend to AI research itself and speed up the pace of technological progress even further.

A World Where One Scientist Has Thousands of AI Researchers

• Clark explained that economic growth comes partly from the discovery of new ideas and the accumulation of human capital.
• But producing a single researcher takes a long time. Training experts in some fields can take 10 to 15 years and millions of dollars, for only a small group of researchers.
• He returned to his opening joke. There is only one of him, and it took almost 40 years to build his experience. What would happen if copies of himself could be made to work at the same time, in their hundreds or thousands?
• AI could make this possible in the form of virtual researchers that can be created in large numbers instantly, as long as there is enough computing power.
• Human scientists might therefore become not less important but more capable, because they can oversee vast numbers of AI researchers working at once.
• Clark cited an economic model developed by a team at Anthropic. Under the assumption that AI can perform many kinds of work on its own and is widely adopted, one model predicts that US GDP could reach $44 trillion by 2030.
• He acknowledged that the number sounds exaggerated. But if the economy were to grow to that level, a great many new industries and inventions would be needed to support the expansion.

From Data Centres to Laboratories, the World Needs a New Round of Infrastructure

• Clark said that over the past 10 years he has travelled to meet many governments to urge them to invest in electricity, data centres and the AI supply chain.
• At the time, few policymakers took the warning seriously.
• Today, computing capacity and data centres are increasingly concentrated in the United States, China and the Middle East, while supply chains face constraints and price pressure.
• He sees building computing infrastructure as only the first phase of the change.
• The next phase is bringing AI into the physical world, with AI running scientific experiments and controlling robots in laboratories.
• The world therefore needs to invest in more scientific equipment, robots and laboratories. Otherwise, the ability to run real experiments could become the bottleneck for discovery.
• Clark cited UNESCO's estimate that the world has around 10 million scientists, which is still not enough for the volume of work and scientific equipment that already exists.
• He asked what would happen if 10 million human scientists could direct 1 billion virtual scientists, or more.
• The countries that invest in building this infrastructure and these supply chains today may be the ones that benefit most from AI.

Developing Countries Have a Chance to Leapfrog, but Must Prepare

• Clark sees this shift as creating opportunities for more than the technology superpowers.
• Developed countries with strong scientific ecosystems can use AI to raise the productivity of the researchers they already have.
• Developing countries that supply raw materials or resources in the supply chain can move up to build experimental and research capabilities.
• He proposed that these countries invest in experimental centres where human scientists work alongside AI scientists, which could help them leapfrog technologically without following the old path in full.
• However, infrastructure alone is not enough.
• If AI helps discover a vast number of new drugs but the clinical trial system can handle only a few, the benefits of those discoveries may never materialise.
• Clark therefore called on policymakers to speed up the process of taking research out of the laboratory and into hospitals, factories and real-world use.

The Smarter AI Gets, the More the World Must Decide What It Should Do

• After describing the economic opportunity, Clark turned to risk. He stressed that AI that can accelerate scientific discovery can also be used to build biological or cyber weapons.
• He said humanity must make at least three important decisions.

1. What Will We Allow AI to Do?

• Even though AI can do scientific research, that does not mean it should be left to do everything without human oversight.
• High-risk biological research in particular requires strict oversight and transparency.
• AI research itself is another area that calls for caution. Letting AI help develop the next generation of AI could speed up progress considerably, but it also raises the risk of losing control.
• Clark said labs such as Anthropic will need to decide how much oversight to apply to AI-driven research, and that rules will likely be needed to govern them too.
• He gave the example of a physics student in Tokyo who asked which experiments humans should do themselves and which should be left to AI. He admitted there is no clear answer yet.

2. Do Humans Need to Understand Everything AI Does?

• Clark said AI is now beginning to produce mathematical work that humans need time to study in order to understand the reasoning and the results.
• Within a few years, AI may be able to design scientific instruments itself, built to suit the way AI works rather than the way humans think.
• The key question is how much humans need to understand of all those tools and processes, and whether AI should be allowed to do work that humans cannot explain.

3. Are We Bold Enough to Imagine a World Very Different From Today's?

• Clark recalled that as a child he worried the 21st century would be filled with crises of population, resources and the environment, so that humanity might not be able to sustain itself.
• Over time, however, he saw technology change constraints that once seemed impossible, from the rapidly falling price of solar power to the development of medicines.
• He also cited things he once never imagined would happen, such as using AI to decode animal sounds and rockets that can return to land on Earth.
• These examples make him believe AI could help humans solve many more problems for which no solution is yet in sight.

The Most Important Technology Humanity Has Built, and One That Demands Restraint

• Clark sees AI as the most important technology humanity has ever built, one that could help address problems that threaten humanity's future.
• At the same time, risk is an inseparable part of the technology.
• He called on governments, scientists and AI companies to work together on serious rules, including oversight of the companies that develop the technology.
• The world should move quickly to build the infrastructure to benefit from AI, but it must develop AI's capabilities carefully and control the pace of progress.
• The central principle is to distinguish between what AI can do and what humans should allow AI to do.
• Clark closed by citing Carl Sagan, in the book Pale Blue Dot, who imagined future humans who are more capable, far-sighted and prudent than humans today.
• He said those are the qualities AI is demanding of humanity, including himself and the company he represents.
• But where Sagan imagined that humans would have hundreds of thousands of years to make themselves wiser and more prudent, Clark said we have a handful of years, less than a decade, easily.

In the end, what stands out from this lecture may not be only the prediction of how much better AI can get. It is the question Jack Clark left with the people shaping the global economy. How can we prepare the infrastructure to benefit from fast-moving technology? And can we build the rules, and the capacity for restraint, quickly enough to keep pace with that progress?

Source

• Jack Clark, Per Jacobsson Foundation Lecture, 2026 IMF-World Bank Annual Meetings, Bangkok (17 October 2026). 

Banga listed clean air, clean water, health care, skilling and the right kind of education as essential foundations. But when people reach working age, they also need the chance to earn, contribute and see a future for themselves.

The World Bank is putting jobs at the centre of its development agenda because the demographic pressure is enormous. Young people in developing economies need work at a time when public finances are constrained, debt is high, aid budgets are under pressure and energy and fertiliser costs have risen.

Three pools of capital, with different jobs to do

Banga described three sources of finance for job creation. Public and concessional money can build the foundations: schools, hospitals, electricity, water and good governance. Philanthropic capital can test new ideas that are still too risky for commercial funding. Private capital can scale approaches that have proved effective, building businesses and employment.

Governments, in this model, create the enabling environment and set the rules. The private sector creates the jobs. Banga put the private sector's share at about 90% of jobs globally, from microbusinesses to large companies.

The World Bank has increased its mobilisation of private investment to US$112 billion in the past year, three times the level of four years earlier. More than half was directed to sectors with significant job-creation potential.

The distribution remains uneven. Private money tends to go where risk is lower. Investment has increased in Africa, Bangladesh, Angola, Argentina, India and Gabon, but the poorest economies have seen little movement. These are often the places where job creation is most urgently needed.

Jobs require power, health and access to opportunity

Banga's argument is that job creation begins before a job exists. People need infrastructure, health, education and skills. The World Bank highlighted several programmes intended to build those conditions.

Through Mission 300, run with the African Development Bank, the Bank aims to connect 300 million people in Africa to electricity. It says 59 million people have been connected so far. It also reported that more than 600 million people have gained better access to health services close to home, against a target of 1.5 billion by 2030. AgriConnect has reached 26 million farmers, while 17 million women and women-owned businesses have gained access to finance and services, and 74 million women now have broadband access.

The institution is also trying to move faster. It says average project preparation time has fallen from 19 months to nine months. The number of performance indicators has been cut from 153 to 22, with greater focus on whether projects improve people's lives. About half of public sector financing is now tied to results, up from roughly a third.

Is mental health part of infrastructure?

The townhall also pushed the conversation beyond finance. One participant asked whether mental wellness and resilience should be treated as part of critical infrastructure, so that they can receive appropriate funding.

Anna Bjerde responded that mental health is already part of the World Bank's health work, particularly in conflict-affected settings where children face prolonged trauma. Paschal Donohoe added that the Bank's education programmes in schools can also help build resilience early in life.

In his opening remarks, Banga set out a test: what did we change, who did we reach, and did it create greater opportunity? The ultimate test, he said, is impact, because if people's lives are not changed for the better, none of it matters.

That was the proposition behind "Small AI, More and Better Jobs," the opening session of the IMF-World Bank Annual Meetings in Bangkok. World Bank Group President Ajay Banga, Singapore President Tharman Shanmugaratnam and IMF Managing Director Kristalina Georgieva discussed AI less as a distant frontier and more as a practical tool for jobs, productivity and access to opportunity.

The backdrop is demographic. Over the next 15 years, 1.2 billion young people will turn 18. If they grow up with clean air, water, health care, education and skills, Banga said, that could be a major demographic dividend. But they will also need work: not only income, but the dignity and hope that come from contributing to society.

AI that works on a 3G phone

Banga contrasted two AI conversations. One focuses on large language models and the possibility that they displace workers. The other asks how AI can make daily work more productive for teachers, health workers and smallholder farmers.

The World Bank calls this second approach "Small AI." The aim is to create tools that are simple, locally relevant and accessible even on a phone connected to a 3G network. Edge AI, which processes information closer to where it is used, could be especially important where infrastructure is constrained.

He illustrated the idea with two scenarios. In a rural Indonesian clinic, a nurse could photograph a patient's skin rash and use a locally developed AI tool to help assess the condition. If it appears to be eczema, the nurse could provide treatment. If the case needs specialist care, the patient could be referred to a regional hospital with an appointment already arranged.

In India, Banga described a possible service for a farmer who cannot read. By photographing the underside of a leaf, she could use a mobile tool to identify a crop disease and receive advice on an appropriate pesticide. The wider platform could help farmers and cooperatives access fertiliser, seeds, crop insurance, credit and better prices.

The point is not to replace a nurse or farmer. It is to connect them to knowledge, services and markets that may otherwise be out of reach.

Five sectors where jobs can grow locally

The World Bank's Jobs Council has identified five sectors with potential to generate employment in developing economies without waiting for outsourced work from richer countries: infrastructure, smallholder agriculture, primary health care, tourism and value-added manufacturing.

Infrastructure includes not only roads, bridges, airports, electricity, water and digital networks, but also skills and education. Smallholder agriculture can become more productive when farmers have access to cooperatives and technology. Expanding community health services creates work for nurses, midwives and technicians while AI can connect them with specialists. Tourism can create employment across many connected businesses. Countries rich in minerals and metals can use value-added manufacturing to create better-quality domestic jobs rather than exporting only raw materials.

Banga argued that governments must build the conditions for investment, while private business is the principal creator of jobs. He estimated that the private sector creates 75% to 85% of jobs in countries. Public policy, infrastructure and risk-reduction tools determine whether private investment can reach the places that need it.

Growth no longer guarantees enough good jobs

Tharman warned that economic growth and job creation are beginning to move in different directions. China, India and several developing Southeast Asian economies have continued to grow, but employment is not necessarily growing at the same pace. AI could widen that gap if governments and companies do not adapt.

He also warned of a K-shaped labour market in which some people and firms gain while others are left behind. One especially urgent concern is the shrinking first rung of the career ladder. If graduates cannot secure a first job, they lose the experience and confidence that make later opportunities possible, while the next cohort of graduates enters the same competition.

His response is to make work-based learning more central to education. Universities need stronger links to internships, real work and dual-learning models. There is no single template because some occupations need deep specialisation while others require people who can combine disciplines and work effectively with AI.

From AI investment to AI adoption

Georgieva described the first economic wave of AI as massive investment in models and data centres. That spending has already supported demand and growth in countries that build AI and in Asian economies that supply the infrastructure. The next wave, she said, is broad adoption.

For developing economies, adoption depends on electricity, digital infrastructure, people's readiness to use technology and public safeguards. The IMF estimates that AI could add around 0.5 to 1 percentage point to growth in advanced Asian economies and 0.2 to 0.8 points in emerging Asian economies. In ageing societies, a one-point gain could be large enough to offset some of the economic drag from demographic change.

The opportunity comes with labour-market risk. The IMF estimates that roughly 40% of jobs globally could be affected by AI over time, including through augmentation, reduced importance of some tasks or replacement. The share is estimated at around 60% in advanced economies and 23% in low-income countries. A lower figure does not mean low-income countries are safe. It can also mean AI has not yet reached them widely.

Georgieva's policy priorities are to track labour-market change, reform education for lifelong learning and reconsider fiscal systems if economies rely less on human labour and more on capital. No one can say exactly what work will look like in five years. Governments can still collect evidence, learn from deployment and adapt.

The leaders' shared aim was not AI for its own sake. It was technology that expands capability: helping people do useful work, giving young people a route into employment and ensuring that the next wave of productivity is not limited to the countries and companies that can afford the biggest models.

• Series A & Financing: Salary Hero has raised US$7.6 million in Series A equity funding led by Cento Ventures, alongside an expanded credit facility with regional banks and lenders to US$10 million, strengthening its capacity to fund on-demand pay across the region.
• Proven Traction: Working with 100+ companies and reaching more than 300,000 employees in Thailand, Salary Hero has helped workers avoid an estimated US$150 million in predatory loans.
• Regional Expansion: The round marks Salary Hero's official entry into Malaysia, supported by Kairous Capital and a dedicated local team.

The Salary Hero Leadership Team (from left): Jonathan Nohr, Co-founder and Managing Director; Tanya Sertthin, Product Lead; Maninthorn Thabsuwan, Head of Solutions; Panthanat Ngamjindasakul, Director of Customer Success; Thanita Uparanukraw, Chief of Staff; Juneray Casuga, Technology Director (front right); Prabhav Rakhra, Co-founder and Managing Director (back right)

BANGKOK & KUALA LUMPUR, 14 October 2026 – Salary Hero, one of Southeast Asia's fastest-growing financial health platforms, today announced it has raised US$7.6 million in Series A equity funding, led by Cento Ventures with participation from Malaysia-based Kairous Capital and M Venture Partners.

Alongside the equity round, Salary Hero has expanded its credit facility with regional lenders to US$10 million. The expanded facility underscores the strength of Salary Hero's repayment infrastructure and the depth of its employer integrations.

The round also marks the company's entry into the Malaysian market.

 We started Salary Hero because the financial system in Southeast Asia is broken for the people who can least afford it," said Prabhav Rakhra, Co-founder and Managing Director of Salary Hero. "A factory worker in Thailand shouldn't have to pay over 20% a month to cover a mid-month gap. We've proven in Thailand that there's a better model: one that's fairer for workers and commercially compelling for employers. Malaysia is the next step.

Fair pay access, embedded in the workplace

 Across Southeast Asia, most working people live paycheck to paycheck. When money runs out before payday, the fallback is too often an informal lender charging punishing rates — sometimes as high as 30% a month. It is a cycle that traps millions of workers in debt.

Salary Hero gives employees on-demand access to wages they have already earned. A mid-month cash gap no longer means a trip to an informal lender. The platform also offers more than 50 financial education classes, budgeting tools, and a growing suite of financial health products designed to help workers take lasting control of their finances.

For employers, the integration is seamless. Salary Hero connects directly to existing HR and payroll systems through scalable integrations with more than 25 enterprise HR partners, so companies can offer the benefit without adding operational overhead. The platform is certified to the ISO 27001 international standard for information security.

The results show up in retention. Companies using Salary Hero, spanning manufacturing, logistics, hospitality, and retail, have reduced frontline worker turnover by 15-25%, translating directly into lower recruitment and training costs. 

A proven engine in Thailand

 Since launch, Salary Hero has built a leading position in Thailand. The platform now reaches more than 300,000 employees across over 100 companies, from large enterprise employers to fast-scaling operators.

 The number that matters most, though, sits outside any balance sheet: an estimated US$150 million in predatory loans avoided: money that would have gone to informal lenders, now staying in workers' pockets.

 Salary Hero sits at an intersection we know well: a financial product embedded into an operational workflow, solving a problem that traditional institutions have ignored," said Ali Fancy, partner at Cento Ventures. "Before venture capital, I spent over a decade running factories in Thailand and saw first-hand what a mid-month cash gap does to a worker and their families. Prabhav and Jonathan have built a solution that fits naturally into the workplace, with impactful benefits to both employees and companies.

AI that predicts turnover before it happens

 For employers in Thailand, turnover is one of the costliest problems, and one the hardest to see it coming. Deloitte's Thailand Salary Policies Survey 2025 found voluntary resignations averaging 12.9% nationwide, climbing to 32.9% in retail. Every departure means more spent on recruitment and training, and weeks of lost productivity while a replacement gets up to speed. 

 Salary Hero is built directly into its partners HR systems, which gives it a view of the workforce that most employers cannot assemble themselves. Working from anonymised partner data, its AI tools show HR teams where turnover risk is concentrated, across sites, teams, shifts, and tenure, as well as how that picture is shifting month on month. 

 Salary Hero then works alongside each partner to act on it. Where the analysis shows people leaving in volume around a particular point in their tenure, a targeted retention incentive at the moment can move the curve for a fraction of what replacing them costs. For HR teams, the value is seeing the problem early enough to do something about it, and putting a retention budget where it earns its keep. 

Launching in Malaysia

The problem Salary Hero solves is just as pressing in Malaysia. According to Bank Negara Malaysia's 2024 Financial Capability and Inclusion Demand Side Survey, 61% of Malaysians struggled to raise RM1,000 for an emergency, up from 47% in 2021. When money runs out, too many workers still turn to unlicensed moneylenders, known locally as "Ah Longs", whose collection tactics can turn violent. Malaysia's Home Ministry recorded 1,965 cases of Ah Long-related violence nationwide between 2020 and 2024.

 The funding powers Salary Hero's expansion into Malaysia with a dedicated local team and partnerships already forming with enterprise clients and distribution partners. Salary Hero began operations in Malaysia in Q2 2026 and is on track to work with 50 corporate partners by the end of 2027. 

 The company enters the market as a proven international operator, combining enterprise-grade technology with strong local networks to reach Malaysian workers at scale.

 Malaysian workers face the same mid-month squeeze we see in Thailand, and too often the same predatory lenders and 'Ah Longs' waiting to exploit it," said Jonathan Nohr, Co-founder and Managing Director of Salary Hero. "That's exactly the problem we've shown we can solve, so we know we can make a real impact here. We're also not starting from scratch: our partnership network and our existing clients across the region have a strong presence in Malaysia, which gives us a running start.

We looked hard at Malaysia’s financial well-being gap and found no one solving it at scale for the workers who need financial access the most, especially the B40 households," said Joseph Lee, Managing Partner at Kairous Capital. "Salary Hero has already proven this model works in Thailand, and our role is to support their expansion to bring a similar playbook into Malaysia, ensuring the entry lands fast and executes right.

KASIKORNBANK (KBank), in collaboration with Techsauce, a technology and business media platform, organized “KBank x Techsauce: EMPOWER THE FUTURE ECONOMY” as an affiliated program held in conjunction with the IMF–World Bank Group Annual Meetings 2026, hosted by Thailand. The event provided a platform for leaders from the public sector, financial industry, and leading global organizations to exchange perspectives on driving the future economy by connecting capital, technology, and human capital development. The event featured a special keynote address by Mrs. Suphajee Suthumpun, Deputy Prime Minister and Minister of Commerce, alongside discussions with senior executives from the International Finance Corporation (IFC), a member of the World Bank Group, Standard Chartered Bank, Microsoft, and Accenture. The event took place at C asean Rama IV, ThaiBev Quarter.

Mrs. Suphajee Suthumpun, Deputy Prime Minister and Minister of Commerce, delivered a special keynote address titled “Reimagining Trade for the Future: Unlocking Growth in a Changing Global Economy,” stating that amid shifts in the global economy, Thailand must rethink its approach to trade and turn change into opportunities for growth. She outlined four key approaches, likening them to four essential keys: rethinking trade partnerships, redefining the value of goods and services, connecting trade with capital, technology, and people, and building trust and resilience. These approaches aim to open new markets, create greater value, and foster inclusive and sustainable growth, particularly by expanding opportunities for Thai businesses and SMEs to access global markets. She also invited Thai and international private-sector partners to work with Thailand as a trusted partner. Emphasizing the importance of collaboration, she noted that no country holds all the keys on its own. Thailand holds some of them, including its strategic location, market, and ability to partner with all sides, while private-sector players and international partners hold many of the others, including capital, technology, and global networks. The door to growth can only be opened when we turn these keys together.

Ms. Kattiya Indaravijaya, Chief Executive Officer of KASIKORNBANK, said, “The global economic landscape is being reshaped by the reorganization of trade, AI’s transformation of how businesses create value, and sustainability’s growing role in investment decisions. We believe Thailand has the potential to be an integral part of the solution, with its manufacturing and services base, business networks, and strategic location connecting it to markets across ASEAN. The future economy is not something we simply wait for; it is something we can all help build. KBank is committed to being your partner throughout this journey, working together to turn opportunities into sustainable growth.”

At the panel discussion, “Powering the Future Economy – Growth, Technology, and Sustainability,” executives from four leading global organizations shared their perspectives. Mr. Sarvesh Suri Regional Vice President for Asia and the Pacific, IFC (World Bank) emphasized mobilizing private capital to support sustainable growth. Mr. Jan Metzger Global Head of Coverage Banking, Standard Chartered Bank highlighted opportunities to connect businesses and capital with global markets. Mr. Azar Koulibaly Regional Vice President and Associate General Counsel, Corporate, External and Legal Affairs, Microsoft Asia focused on expanding access to AI and developing digital skills. Mr. Anoop Sagoo CEO, Southeast Asia, Accenture discussed leveraging AI to enhance business efficiency and competitiveness.

This event aimed to bridge perspectives from the public sector, the financial sector, the business community, international organizations, and global technology leaders to collectively transform global challenges into growth opportunities for Thailand and build the future economy with sustainable competitiveness.

AI can answer a cybersecurity question in seconds, but judging whether that answer is right still takes someone who understands the work deeply. And if AI absorbs all the entry-level work, a harder question follows: where will the next generation gain the experience to become that kind of expert?

That question ran through the Workforce Development Panel, "Supercharging the Cyber Workforce: AI, Skills, and Accreditation," at Singapore International Cyber Week (SICW) 2026. The session opened with an address by Tan Kiat How, Singapore's Senior Minister of State, who used the occasion to announce that Singapore will establish a Cybersecurity Professional Accreditation Framework. Lightning talks from industry leaders followed, and the session closed with a panel of UK representatives who shared lessons from the path their country began walking years ago.

Although the speakers came from different corners of the field, they agreed on one point. AI does not make cyber professionals less important. If anything, it raises the bar for what it means to be a professional, and the field now needs a system that can show who has actually cleared that bar.

AI Answers Fast, but Who Can Tell if the Answer Is Right?

Tan Kiat How began with a workforce study by the Cyber Security Agency of Singapore (CSA). Eight in ten respondents said their organizations were already using AI-enabled cyber defense tools, while seven in ten rated human judgment and problem-solving skills as essential to using AI effectively.

These figures suggest that AI is not diminishing the role of people but reshaping it. As AI takes on more tasks, cyber professionals will increasingly need to direct these tools, interpret their outputs, validate their conclusions, and make the decisions that follow. Tan Kiat How noted that AI fluency will become part of cybersecurity competence, but stressed that it "is not the same as cybersecurity expertise." AI can generate an answer quickly, yet knowing whether it is the right answer still requires deep professional knowledge.

A separate CSA study, which surveyed around 1,000 cybersecurity and HR professionals, pointed in the same direction. According to Wong Choon Bong, Director of CSA's Workforce and Strategic Partnership Division, more than three-quarters of the cybersecurity professionals surveyed see AI as more of a career opportunity than a threat. Many, however, also worry that some of their skills may become less relevant. In the AI era, Wong Choon Bong concluded, the profession needs to "do more, not less."

Automate the Task, Not the Learning

While encouraging organizations to embrace AI, Tan Kiat How cautioned that they should think carefully about what they automate. He borrowed a distinction that he and his colleagues in medicine often draw between two kinds of friction. Meaningless friction, such as administrative work, is worth streamlining. Meaningful friction, on the other hand, is how people learn by doing. The principle to follow, he said, is to "automate the task, not the learning."

This matters because decisions in cybersecurity often carry real consequences for organizations and the people they serve, and the ability to make the right call comes partly from trial and error. Tan Kiat How therefore asked a pointed question: if automation absorbs both the task and the learning, how will junior professionals make mistakes in a safe environment and learn from them?

Dr. Ning Wang, CEO of Offensive Security (OffSec), echoed this concern. Leaders, she said, must keep enough junior positions open, because no one can become a senior expert without first going through the journey of a junior, making mistakes and working through the struggles along the way. Wong Choon Bong added that the efficiency organizations gain from AI today should not come at the expense of the human expertise they will need tomorrow.

Cyber Analysts Are Shifting From Doing the Work to Orchestrating It

As AI moves into every stage of security operations, the human role is shifting with it. Dr. Lim Woo Lip, Chief Technology Officer for Cyber at ST Engineering, explained that analysts once did nearly everything by hand, from monitoring systems and reviewing alerts to investigating incidents. Today, they work alongside AI that processes information and accelerates investigations, while humans supervise the outputs and make the final decisions. In the future, the role may evolve into an "AI-augmented security orchestrator," someone who coordinates AI-supported operations, validates findings, and helps shape security strategy.

That shift calls for skills well beyond cybersecurity knowledge. According to Dr. Lim, professionals need to understand where AI performs well and where it may fail, know how to secure and govern AI, and think critically enough to challenge its outputs, because "speed is not equal to accuracy." They must also connect cyber risk to business impact and keep learning as both AI and threats evolve. Employers, for their part, need to redesign roles to decide where AI can help and where human accountability must remain, reskill their workforce, and set clear guardrails for AI use.

Dr. Ning added another layer. Professionals, she argued, should focus on outcomes rather than output, since no amount of activity matters if it fails to solve the problem at hand. And the more AI can do, the more fundamentals matter, because without a real grasp of the subject, no one can judge whether AI got it right.

Singapore Will Define What a "Good" Cyber Professional Looks Like

If the work is changing, the profession must change with it. Professionals need clarity on which capabilities to develop, employers need to understand what good looks like, and training providers need to know what the industry will need next. To that end, Tan Kiat How outlined two major steps. First, CSA will work with the Infocomm Media Development Authority (IMDA) and Singapore's skills and workforce development agency to develop a National Competency Framework for Cybersecurity, which will set competency standards for cybersecurity roles, including the use of AI.

The second step is the accreditation framework itself, which will validate and recognize the knowledge, skills, and experience of cyber professionals. Wong Choon Bong explained that candidates will be assessed on a combination of demonstrated technical competencies, relevant professional experience, qualifications and industry certifications, and professional ethics. This allows experienced professionals to prove their capabilities through practical skills and a professional track record, not just the certifications they have collected. Accreditation committees made up of leading experts from different specializations will help develop the criteria and take part in assessments, because professional standards must reflect real-world practice rather than be set by the government alone.

Singapore plans to take a measured approach. The Cybersecurity Professional Office, established in collaboration with the Association of Information Security Professionals (AiSP), will work with the cybersecurity community to develop the details, with a voluntary pilot targeted for the second quarter of 2027 in selected specializations and participating organizations. "We are not trying to create another layer of certificates," Tan Kiat How emphasized. The goal is a framework that gives employers greater confidence, gives professionals greater recognition, and gives the profession a stronger common standard.

Beyond assessment, Singapore also wants to invest in people at every career stage. For young talent, CSA has just signed an agreement with Trend Micro to organize Singapore Cyber Conquest, the national Capture-the-Flag (CTF) competition, at SICW for the next three years. For senior leaders, Tan Kiat How pointed to "a different kind of fluency": the ability to speak with boards, policymakers, and regulators in terms they understand, since cybersecurity is no longer just a back-office function.

A Certification Shows You Passed. Professional Accreditation Shows You Are Accountable

While Singapore is only at the starting line, the UK has been on this road for years. The second half of the session, moderated by Professor Steven Wong, Advisor to AiSP, opened with the most basic question of all: why did the UK choose to professionalize cybersecurity in the first place?

Felicity Powell, Policy Lead for R&D, Skills, and International at the UK's Department for Culture, Media and Sport (DCMS), explained that as the market rushed to train people to close the skills gap, credential inflation followed, making it hard to tell who was genuinely competent. Professionalization answers that problem because it "tells you what good looks like." Felicity compared it to aviation: passengers trust that their pilot can land the plane safely because a professional framework stands behind that pilot, just as one stands behind doctors, lawyers, and engineers. Cybersecurity, she argued, has become critical infrastructure in exactly the same way.

Giles Grant, CEO of the UK Cyber Security Council, drew a clear line between three things that are often confused. A qualification answers whether a person has been taught. A certification answers whether they can pass a structured technical test. Professional registration, however, answers whether they are accountable. It comes with a code of ethics, continuing professional development (CPD), periodic reassessment, a public register from which titles can be revoked, and peer assessment.

Scott Beale, CEO of ISC2, added that passing an exam only proves what someone knew on a given day, whereas belonging to a system that requires continuing education shows that they are keeping pace with a fast-changing field. And because cyber threats routinely cross borders, a common language of knowledge and ethics helps organizations trust the people they work with, wherever those people are based.

How Do You Assess Judgment When an Exam Cannot?

Knowledge and skills can be tested in an exam, but ethics and judgment are far harder to measure. Debi McCormack, Programs and Partnerships Director at The Cyber Scheme, one of the bodies licensed to assess candidates, said one surprise was how much guidance candidates needed. Technical professionals are so used to working in teams that they instinctively say "we did this." Assessors had to draw them out until they could say "I made this decision." That matters most in ethical situations, such as when a client asks for a finding to be left out of a report. Explaining that you decided to refuse demonstrates both your ethics and your ability to make a judgment call.

On assessment design, Nick Benson, CEO of CREST International, said the challenge is to be "rigorous and workable." His first principle is that candidates should not have to go through the same assessment twice when they have already passed one that is at least as rigorous. His second is to put employers at the center, since they see their employees' conduct and ethics in real work every day. His third is to allow the same standard to be met in different ways, accommodating differences in location and learning style, including candidates who are neurodivergent.

Lessons From the UK: Don't Start at the Top of the Pyramid

The question Singapore most wanted answered was which mistakes to avoid from the outset. Giles acknowledged that the UK took years and initially got it wrong, because it started with process rather than with demand. The hardest part, he said, is not building the framework but winning hearts and minds, especially among professionals who already suffer from certification fatigue. His advice: "Think big, start narrow. Test, verify and scale." He added that if the Singapore government requires accreditation for certain public sector roles, it would go a long way toward driving early adoption.

Debi added that the UK's biggest misstep was to begin by assessing at the Chartered level, the very pinnacle of the profession. As a result, most professionals felt the titles were irrelevant to them and out of reach. Her advice is to start with the core of the workforce and to listen closely to the first candidates who go through the process, because if it feels unwieldy or unfair, no one will sign up.

In the UK, demand has grown fastest where titles are mandated, such as work on critical national infrastructure and in the public sector. Still, Debi has noticed that candidates' initial reluctance tends to turn into pride once they earn their titles, and over the past 12 months, more people have been saying "I chose to do this" rather than "my boss told me to." On cost, which remains a barrier for young professionals, Debi and Nick agreed that employers need to invest more, much as law firms invest in the careers of trainee lawyers, because employers ultimately benefit from having qualified people on their teams.

As AI Raises the Bar, Professionals Must Rise With It

Both halves of the session arrived at the same place. In closing, Tan Kiat How said that while AI will transform many cybersecurity tasks, it will also raise the premium on expertise, judgment, and responsibility. The goal, therefore, is broader than training people to use the latest tools. It is to build professionals who understand technology, know its limits, and can make sound decisions when it counts.

When AI can answer almost any question, the most valuable asset becomes the person who can tell which answer is right, along with a system that can confirm that person deserves to be trusted.

Source: Workforce Development Panel, "Supercharging the Cyber Workforce: AI, Skills, and Accreditation," Singapore International Cyber Week (SICW) 2026 (Opening Address by Tan Kiat How; lightning talks by Dr. Ning Wang, Dr. Lim Woo Lip, and Wong Choon Bong; and the panel discussion "UK's Professionalisation Journey, and Lessons for Singapore's Implementation")

In the depths of winter in December 2025, Poland was hit by a coordinated wave of cyberattacks. The targets included more than 30 wind and solar installations, as well as a combined heat and power plant that supplied heat to nearly 500,000 people. Months earlier, in April 2025, South Korea's largest mobile carrier had reported a breach that compromised sensitive data from nearly 27 million user records.

Were these isolated accidents, or pieces of a larger picture that no one has yet seen in full?

That question set the tone for the Opening Ceremony of Singapore International Cyber Week (SICW) 2026, held under the theme "Cyber at a Crossroads: Shaping the New Cyber Order." Three speakers took the stage: 

• Gwenda Fong, Chief Executive of the Cyber Security Agency of Singapore (CSA) and Commissioner of Cybersecurity
• K Shanmugam, Senior Minister, Coordinating Minister for National Security, and Minister for Home Affairs of Singapore, who used the occasion to launch the Singapore Cybersecurity Strategy 2026
• Amandeep Singh Gill, UN Under-Secretary-General and Special Envoy for Digital and Emerging Technologies.

Each spoke from a different vantage point, yet their arguments converged on a shared conclusion. No one can see the full picture of cyber threats, and as AI makes attackers faster, defenders can no longer afford to wait for the pieces to fall into place on their own.

Everyone Sees Something, but No One Sees Everything

Fong, delivering her welcome remarks at her first SICW as CSA Chief Executive, began with Singapore's position in the world. As a small country whose success was built on being open and connected, Singapore depends on digital infrastructure, technology platforms, and supply chains that extend far beyond its shores. That connectivity has brought enormous opportunities, but it also exposes the country to risks that originate beyond its borders.

The difficulty is that these threats span multiple systems, organizations, and countries. Governments, infrastructure operators, cloud providers, and cybersecurity companies each have visibility of a different part of the threat landscape. As Fong put it, "Each sees something important but no one sees everything."

Sophisticated attackers exploit precisely this gap. They operate across multiple systems and networks, often over extended periods, so what looks like an isolated incident to one organization may in fact be part of a much larger campaign. If your organization spotted a minor anomaly in its systems today, how would you know whether it was a one-off event or just one corner of a much bigger picture?

Fong added that assembling the picture is only half the challenge. The other half is doing it quickly enough to act before the adversary does.

When Hiring More Analysts Cannot Keep Pace, What Do Defenders Have Left?

Speed is exactly where AI changes the equation. Singapore has long relied on technology to overcome its constraints of size and manpower, and it intends to harness AI in the same way. Because the country cannot simply keep adding analysts as threats grow in volume and sophistication, Fong said, it needs AI to help its cyber defenders "to see more, to understand faster and to act more effectively."

Attackers, however, can harness AI too. In Fong's view, the real challenge is to use AI to strengthen defenses without allowing it to tilt the advantage toward the attackers.

AI is also becoming part of the environment that must be secured. As AI agents gain access to data, applications, and tools, they create new dependencies and vulnerabilities. For years, Fong noted, cybersecurity has required an understanding of what is connected to what. In an increasingly agentic world, defenders must also understand "what can act on what, with what permissions and with what consequences."

This warning is far from theoretical. Speaking at the same ceremony, Shanmugam said that AI agents can be manipulated into carrying out unintended actions that compromise essential systems, and that agentic AI is advancing faster than safeguards can keep up. AI agents have already carried out unauthorized tasks and accessed systems they were never meant to reach. Just last month, Australia disclosed that an AI agent had gained unauthorized access to non-public data on a government portal.

Can your organization say, right now, which systems each of its AI agents is allowed to touch?

Waiting for a Breach Before Responding Is a Mindset Singapore Is Leaving Behind

For Shanmugam, AI-enabled cyber threats are "really not a matter of if, but when." Singapore cannot resist AI adoption, so it must find ways to adopt AI safely and securely. Nor are recent attacks isolated incidents. Advanced Persistent Threat (APT) groups have both the intent and the capability to target critical infrastructure persistently, for espionage as well as disruption, and every country must plan on facing such attacks with increasing frequency.

Singapore has learned this lesson firsthand. At last year's SICW opening, Shanmugam revealed that UNC3886, an APT group, was attacking the country's critical infrastructure even as he spoke, having used sophisticated tools to gain access to its telecom networks. Singapore responded with Operation Cyber Guardian, its largest cyber response to date, which brought cyber defenders from the government together with telcos and other partners.

That experience shaped the new strategy, under which Singapore will seek to "act earlier, act at national scale, and act at the speed of the threat." 

The strategy is also broader in scope than the way national cyber defense has traditionally been viewed. It covers strengthening critical infrastructure and the wider cyberspace, building capabilities in AI and quantum, deepening global partnerships, and investing in the cybersecurity ecosystem and workforce.

The first principle is to act earlier. Traditional cyber defense relies heavily on protecting systems, conducting periodic assessments, and responding after something has gone wrong. Adversaries, however, are continuously looking for vulnerabilities and trying to establish persistent access, so defense must be continuous as well. That means monitoring and hunting for threats on an ongoing basis, identifying weaknesses before adversaries do, and fixing them quickly. If attackers are scanning for vulnerabilities every day, do periodic system checks still count as protection?

Attackers Don't Treat Government and Industry as Separate. Why Should Defenders?

The second principle, acting at national scale, comes directly from the lessons of Operation Cyber Guardian. Shanmugam explained that sophisticated attackers do not see the government and telcos as separate silos. They look across all systems for the weakest point. National defense must therefore look across the whole system as well, with government and industry working closely together in any response.

To that end, Singapore has established the National Cybersecurity Command Centre (NC3) as CSA's operations hub for national cyber defense. NC3 serves three functions. First, it provides a national picture of both vulnerabilities and threats by supporting the continuous monitoring of critical infrastructure systems and by bringing together threat information from the government, Critical Information Infrastructure (CII) owners, cybersecurity companies, and international partners. Second, it helps detect APT groups earlier through proprietary threat detection systems and privileged intelligence, allowing defenders to act before attackers can cause serious harm. Third, it enables an effective national response to major cyber threats, similar to Operation Cyber Guardian, through AI-enabled coordination systems.

The government will also provide greater protection upstream by developing a National Protective Domain Name Service (PDNS). When someone tries to connect to a domain known to be malicious, PDNS can stop the connection before the threat reaches the user, reducing exposure to phishing, scams, malware, and other cyber threats.

If Attackers Use AI, Defenders Cannot Afford to Be Slower

The third principle is to act at the speed of the threat. Shanmugam's logic is straightforward: if AI allows attackers to identify vulnerabilities and launch attacks more quickly, defenders must use AI too. The government will progressively deploy AI across its national cyber defense operations, including within NC3 and the Government Cyber Security Operations Centre, and will work with industry and technology partners to speed up the adoption of AI-enabled cyber defense across critical sectors.

Some of these capabilities are already in place. CSA has developed AI tools that automate the discovery and mapping of exposed assets and vulnerabilities, and that help cybersecurity specialists test complex vulnerabilities in web applications. Meanwhile, the Government Technology Agency (GovTech) uses AI agents to accelerate vulnerability discovery and security testing.

That said, Singapore does not plan to build its own AI models for every use case. What it really needs, Shanmugam said, is the capability to evaluate the best available models and adapt them to its own needs. The government will therefore invest in AI harness engineering, which will give it the flexibility to choose between AI models and to securely integrate the best ones with the tools, data, and workflows that cyber defenders need.

Trust Is Not a Soft Concept. It Is Infrastructure

Even the fastest tools mean little, however, if no one trusts the systems in which AI operates. Gill tied this point to the conference theme. Until now, he observed, we have treated cybersecurity measures like guardrails along a road. As AI accelerates, though, the road itself and the vehicle we are traveling in are both shifting, which means the crossroads is shifting too. Our traditional ways of looking at threats and responses must therefore change radically.

At the root of the problem lies a dilemma. To unleash the potential of AI, agents must be given more autonomy to act in the physical world. That expands the attack surface available to hackers and increases the likelihood of mistakes in how tasks are executed. Both digital and physical systems must therefore become more secure as AI adoption moves forward.

Gill argued that "in this AI world, trust should be seen as hard infrastructure" rather than a soft concept. We trust a digital system when we understand what it is designed to do, when we can test what it claims to do, when responsibility is clear if it fails, and when humans retain the ability to intervene. Trust therefore cannot rest on confidence in the technology alone. It must be earned through compelling evidence across domains and across borders.

This kind of trust matters to citizens and governments alike. According to Gill, public trust is one of the biggest obstacles to AI adoption today, while governments must decide whether to deploy AI systems in critical sectors and whether they can rely on systems developed in another country. "We do not partner those we do not trust," he said. "And we cannot trust what we cannot see clearly, hold to account or govern."

The question becomes even harder when a model is developed in one jurisdiction, trained on infrastructure in another, integrated by providers elsewhere, and deployed simultaneously across many societies. When a system like that fails, who is responsible?

Diversity Is a Strength. Fragmentation Is the Risk

The answer does not lie in requiring every country to govern AI the same way. Gill noted that at the UN General Assembly, more than 110 countries, including Singapore, referenced AI in their speeches. Even within ASEAN, approaches vary widely, from Vietnam's law-based model to Singapore's approach, which rests on strong principles, concrete guidance, testing tools, and meaningful human accountability.

This diversity is not the problem. The real problem is fragmentation and a retreat into silos. Gill sees diversity as a source of strength, resilience, and experimentation during this early phase of AI development, and as a foundation for greater interoperability across jurisdictions. After all, "a vulnerability that's discovered in one place can become a threat anywhere," and a lesson learned in one place should enhance protection everywhere.

For that reason, the measures companies take on their own, whether or not governments encourage them, will not be enough. Gill called for a basic minimum layer of international cooperation built on three elements. The first is helping every country build a baseline level of safety infrastructure, whether it develops, borrows, or buys its AI models. He pointed to the Singapore AI Safety Institute, based at the Digital Trust Centre, and to the UK's approach as models, since both operate separately from regulation while working with the private sector and keeping policymakers informed.

The second element is minimum standards for testing, verification, the certification of independent auditors, baseline system behavior, and red lines, comparable to accounting standards in the financial sector. The third, and perhaps the most important in Gill's view, is information sharing, including incident reporting, transparency, and even data on near misses, which can offer valuable lessons. Because this cannot be achieved through ad hoc arrangements, he argued, the world needs a multilateral risk management framework backed by independent and credible foresight.

Compete if You Must, but Don't Wall Yourself In

Gill's proposal aligns with a reality that Shanmugam openly acknowledged: even a more proactive defense cannot prevent every attack, which makes cooperation all the more important. He also urged realism about the current environment. Strategic competition will intensify, particularly over critical technologies such as AI, and countries will act to safeguard their national interests. Singapore will do the same.

The danger arises if every country responds to that competition by building inward and acting alone, because doing so would break up the digital ecosystem that everyone depends on. Shanmugam said signs of this are already visible in advanced technology domains such as AI, semiconductors, and telecommunications. Cyber threats cross borders, and technology and infrastructure are spread across jurisdictions, so no country, however capable, can secure the digital domain on its own.

"We will not agree on everything," Shanmugam said. Even so, countries must build practical cooperation to exchange information, respond to incidents, tackle common threats, and reduce the risk of wider instability. That cooperation must also extend beyond governments, since much of the infrastructure is owned and operated by industry, and much of the technology, threat intelligence, and technical expertise sits outside the public sector.

Fong added that cooperation must go beyond agreeing on principles and translate into the ability to work together when it matters. She cited the ASEAN CERT Incident Drill, which since 2006 has brought national cyber response teams together to practice responding to cross-border incidents, building the relationships and habits of cooperation that count in a real crisis.

This year, SICW is applying the same practical approach to AI, moving beyond discussing what AI might mean for cybersecurity toward testing what it means in practice. CSA is partnering with Google on an agentic AI cyber scenario exercise for international partners, working with the AI Safety Hub on a Frontier AI Governance Workshop for officials from ASEAN and the broader Asia-Pacific region, and joining Cloudflare to deliver hands-on training in AI-enabled cyber defense for critical infrastructure operators. As Fong put it, some things "can only be learned by working through the problems together."

Resilience Does Not Mean Closing the Door

All of this leads back to the question facing a small country like Singapore: how can it protect itself without shutting out the world? Shanmugam's answer is that Singapore will build its resilience by strengthening its own capabilities and safeguarding its national interests, but "resilience does not mean retreating from the world." On the contrary, resilience is what gives Singapore the confidence to remain connected. That, he said, is ultimately what the Cybersecurity Strategy 2026 is about: protecting the trust on which the country's digital economy and society depend.

Fong closed her welcome remarks by encouraging everyone to look beyond their own vantage point and to "find the people in the room who can help us see the rest."

If no one can see the whole picture of cyber threats, the question every organization must answer is not what it can see, but who it is looking with.

Source: Opening Ceremony of Singapore International Cyber Week (SICW) 2026 (Welcome Remarks by Gwenda Fong, Opening Address by K Shanmugam, and Special Address by Amandeep Singh Gill), CERT Polska

Conversations about national cybersecurity tend to conjure up images of expensive defense systems, massive budgets, and security agencies working behind the scenes. Liisa-Ly Pakosta, Estonia’s Minister of Justice and Digital Affairs, started somewhere far less technical: with trust.

At Singapore International Cyber Week (SICW) 2026, held under the theme “Cyber at a Crossroads: Shaping the New Cyber Order,” Pakosta delivered a keynote speech before joining a fireside chat with Gwenda Fong, Chief Executive of the Cyber Security Agency of Singapore (CSA) and Singapore’s Commissioner of Cybersecurity. The conversation began from a challenge both countries share: as small, highly digitalized nations, they need cyberspace to remain “open, secure, and trusted.”

That common ground surfaced early in the keynote. Pakosta noted that Estonia, like Singapore, ranks among the most digitalized countries in the world, and compared the origins of Estonia’s digital state to the way Singapore solved its water problem. Both, according to the minister, grew out of an urgent need to engineer a solution that was “great, but cheap and extremely effective.” Pakosta even suggested that Singaporeans might think of Estonia every time they turn on a tap.

A Country With No Money Had to Build Its State on Trust

Estonia’s urgent problem arose in the 1990s, after the country regained its independence from Soviet occupation. It had almost no money, yet it wanted to build an effective democratic system of governance as quickly as possible.

According to Pakosta, the engineering behind Estonia’s digital state rests on three principles, all of which lead back to trust. The first is cybersecurity. The second is full transparency, which is why many public services are built on open-source software, including the country’s online voting system. Pakosta described e-voting as perhaps the most scrutinized and most targeted system in the world, yet because most of its code is open, Estonia and its partners have found vulnerabilities well before malicious actors could.

The third principle is close cooperation with the private sector. In Pakosta’s view, innovation comes from companies, since bureaucrats are good at many things but rarely at innovating. By taking on that risk and letting in the brightest ideas, the Estonian government has digitized almost all of its public services. The one exception the minister cited was divorce. (Couples can now file for divorce online, but they still need to meet an official in person once, according to e-Estonia.)

With databases this tightly connected, the state can deliver services before citizens even ask for them. Since 2014, Estonia has offered life event-based services. When a child is born, for example, families who have given prior consent automatically receive everything they need, from a family doctor and child benefits to the child’s documents. Pakosta stressed that this works only because people own their data and trust the government enough to share it, which in turn demands an extremely high level of cybersecurity.

The 2007 Lesson: Trust Must Be Protected Together

A state whose data is entirely digital, interoperable, and high-quality also becomes a high-value target. Pakosta said that in 2007, Estonia became the first country ever to be hit by a cyberattack from another state, an attack the minister attributed to Russia. At the time, Estonia could still respond by switching off the internet, an option that no longer exists today.

Estonia already had strict cybersecurity rules before 2007, but after the attack, every rule became mandatory and cooperation with companies grew much tighter. The reasoning was simple: when every system is interconnected, it matters little whether a breach happens on the corporate side or the government side, because the damage falls on the same shared trust.

That cooperation extends beyond companies. Pakosta explained that Estonia can also call on a large number of volunteers when a serious situation arises. The volunteer network is well organized, and every member undergoes a background check. This support has become even more important, as Estonia has seen a dramatic rise in cyberattacks since the full-scale escalation of the conflict in Ukraine.

Arrows and Shields: When AI Agents Lead the Attack

Even so, Pakosta acknowledged that 100% cybersecurity does not exist. Throughout history, the minister noted, whenever someone invented an arrow, someone else invented a shield. Governments therefore have to keep adapting under what Pakosta called a “new arrow, new shield” policy.

The newest arrow came up when Fong asked which development had most changed the minister’s thinking about cybersecurity. Pakosta pointed to the moment it became clear that AI agents could lead cyberattacks on their own, comparing the feeling to seeing artillery for the first time after a lifetime of arrows and shields. The leap in scope and speed is so great that the old approach, in which humans gather to work out the best response, is no longer fast enough.

As a result, Estonia has shifted from a reaction model to a continuous monitoring model, because there is no longer any time to prepare a response. Defense has effectively become AI versus AI, with humans still moderating. At the same time, the government has centralized much of its organization to keep pace. Pakosta shared these changes in response to an audience question about how the country’s election system is coping with attackers who use agentic AI.

On another front, Pakosta sees the fundamental task as unchanged: protecting sensitive data, documents, and interoperability. The difference now is that interoperability is extending beyond national borders. The minister hopes to enter Singapore soon with a digital passport stored in a digital wallet rather than a paper one, a change Pakosta believes is “not days, but hours” away and one that will require an entirely new level of cybersecurity.

Cyber Literacy Starts in Kindergarten

Because the challenge is not purely technological, Pakosta emphasized that cybersecurity also has cultural, political, and diplomatic dimensions. On organizational and societal change, the minister drew on experience as a parent of five: children who do not make their beds do not get pocket money. Estonia applies the same logic to cybersecurity, with very concrete rules that everyone is expected to follow.

The bigger shift, however, has happened at the societal level. Estonia teaches digital skills and cybersecurity starting in kindergarten, and the effect reaches far beyond the classroom, because children go home and share what they have learned with their grandparents. As a result, society as a whole has developed a solid understanding of something as invisible as cybersecurity.

The government also offers free cybersecurity and AI training to everyone. Pakosta gave the example of an elderly woman in a remote village who can take both courses at the local library, since the librarians there have already been trained.

Citizens Shouldn’t Have to Know Whether a Service Comes From the State or a Company

From education, the conversation turned to public-private partnership. Fong observed that governments are ultimately accountable for national security, yet much of the digital infrastructure is owned and operated by private companies. Has the balance between the two been struck correctly?

Pakosta replied that the goal of a fully digital society like Estonia’s is for people to barely notice that a government exists, because services simply reach them. The flip side is that people often cannot tell whether a service came from the public or the private sector. The government must therefore work with companies to maintain trust, which matters to both sides well beyond technology or data breaches.

This partnership must be backed by thorough risk assessments that are continually renewed as new risks emerge. Pakosta explained that Estonia defines digital sovereignty in terms of risk rather than geography, which sets it apart from many other countries. The guiding principle, in the minister’s words, is “risk assessment, risk assessment, risk assessment, ongoing risk assessment and no vendor lock.”

A Voice for Small States as Big Tech Grows More Powerful

Looking further ahead, Fong raised threats that may only fully emerge years from now, such as quantum computing, which could break the encryption in use today. Pakosta answered that the one thing humankind can do is make rules. Pointing to nuclear energy, the minister noted that it took from August 1945 until 1970 for the world to regulate it. This time, the world must move much faster, with global rules for both AI and social media.

Here, the similarities between the two countries came into focus again. Pakosta said Estonia and Singapore share the same position on social media: rather than banning it for children, both want platforms to change so that they become safer for everyone, keeping the good and removing the bad.

Fong’s final question went straight to the heart of both countries’ situation. As small, highly digitalized states operating in quite different geopolitical environments, how can they work with industry to secure a meaningful voice in decisions that shape their digital security, especially as big tech grows more influential?

Pakosta said global law and order matters most, along with regulation that prevents the digital divide from widening. Possible solutions include open source, the inclusion of smaller languages and cultures, and greater transparency. In the minister’s view, the growing digital divide is an even worse scenario than cyberattacks led by AI agents. Those attacks can be handled, but a widening divide, both between individuals and between countries, would be a far bigger problem.

Trust Is Hard to Build and Easy to Spoil

Throughout the keynote, the answer Pakosta kept returning to was cooperation among like-minded partners. When a cyberattack hits one place, there is likely spillover somewhere else. Learning from one another is not easy, because it requires openly sharing what happened, which is rarely pleasant. Yet, the minister argued, it is the only way to stay protected. Bad actors may sometimes be more innovative, but whenever good actors unite their ideas, skills, and knowledge, they prove stronger.

Pakosta pointed to Estonia’s cooperation with Singapore as an example, describing the two highly digitalized countries as like-minded partners that support and learn from each other, a relationship that has meant a great deal to Estonia in practice. Fong, in turn, closed the conversation by noting that the future of cybersecurity will be shaped not only by technological advances but also by the choices made today about whether and how to work together.

Pakosta ended on trust itself: hard to build, easy to spoil, and never just a technical issue. It extends to how societies teach their children, support their elderly, make sure no one is left behind, and rebuild confidence after a breach.

For small nations like Estonia and Singapore, the strongest cyber defense may not be technology at all, but the trust that everyone helps protect.

References: Keynote and fireside chat “Liisa-Ly Pakosta x Gwenda Fong,” Singapore International Cyber Week (SICW) 2026; e-Estonia

Can an AI company built in Thailand reach global scale when its home market is far smaller than those of the technology leaders emerging from the United States and China?

At Techsauce Global Summit 2026, Korawad Chearavanont, Founder and Executive Chairman of Amity, offered one answer. In the session "$160M Raised: Building a Global AI Player from Thailand," he described a strategy that links specialized models, acquisitions, financing and regional talent. The aim, he said, is not to recreate OpenAI. It is to build an AI company from Thailand that can compete globally.

That distinction matters. Amity's approach is less about winning a race for the largest general model and more about controlling the data, distribution and execution needed to solve narrower enterprise problems across multiple markets.

Smaller models can win when the data is specific

Korawad argued that the opportunity for companies in the region lies in Vertical AI, meaning models designed for a particular industry, geography or task. Frontier models learn largely from broadly available data. Yet much of the information that matters inside an industry never reaches the public internet. He used data flowing through Thai financial institutions as an example of information that general frontier models cannot access.

That gap creates room for a specialized model. Korawad contrasted frontier models with hundreds of billions or as many as one trillion parameters against vertical models with 7 billion to 10 billion parameters. In his account, a smaller model trained on relevant proprietary data can perform a specific task more effectively and at a lower computing cost.

The competitive advantage, then, does not have to come from building the biggest model. It can come from securing data that larger rivals lack and choosing problems where domain knowledge matters more than general capability.

A small home market demands global ambition

"Our dream and our vision is to build a global AI player out of Thailand, at least a globally competitive player," Korawad said.

Korawad argued that Thailand or Southeast Asia alone would not provide a large enough market base for Amity to achieve the scale and research capacity required for global competition. He said more than 80% of Amity's profit already comes from outside Southeast Asia, with Europe serving as a major geographic focus.

His strategic frame combines Europe and Southeast Asia into a market representing roughly $25 trillion to $26 trillion in GDP. Korawad compared that footprint with the size of China's economy. The figure was central to his argument that a company needs access to a sufficiently large market if it is to fund research and withstand competition from much larger American and Chinese players.

For a technology company originating in a smaller economy, international expansion is therefore more than an additional revenue stream. In Korawad's view, it is part of the infrastructure required to become globally competitive.

Buying distribution instead of rebuilding it country by country

Entering each new market organically means hiring a sales team, adapting the product and language, and establishing customer relationships from the ground up. Korawad said that after 14 years of trying to build a technology company from Thailand, he concluded that this path offered a low probability of success against larger incumbents.

Amity responded by adopting an acquisition-led rollup strategy in 2023. It buys software businesses that already possess customer distribution and proprietary data. Amity's teams then introduce AI into the core product, develop vertical models around the business and offer those capabilities to the acquired company's existing customers.

Each acquisition therefore serves two connected purposes. It provides an established route to market, and it opens access to data that can support specialized models. The acquired software business becomes a distribution channel, a source of domain data and an operating environment in which AI can be deployed against real customer needs.

Korawad noted that disciplined target selection is essential. Amity looks for companies whose products its AI teams can transform quickly and whose data, markets and customer base fit the broader strategy. The work begins after a deal closes, when the expected link between the acquisition and the AI platform has to be made operational.

Debt came before the $160 million in funding

Acquisitions require capital before their results are proven. Korawad said Amity initially found it difficult to raise money for the AI rollup. The company used private credit and debt to finance its first acquisitions, then applied AI to those businesses and sought to demonstrate growth.

Once the model had evidence behind it, Amity raised $60 million in 2024. It repeated the process of acquiring software companies and transforming them with AI, leading to its latest $100 million round. Together, the two rounds brought the total to $160 million.

The sequence is as important as the headline number. Korawad and his team took early financing risk, produced operating evidence and only then secured larger pools of capital to accelerate the approach. In his telling, funding followed proof that the acquisition and transformation model could be repeated.

Global revenue, regional execution

Although most of Amity's profit comes from outside Southeast Asia, Korawad described talent in Thailand and the wider region as an execution advantage. The company's research labs are based in Bangkok and Singapore. He also referred to specialized talent in Europe and a team of a couple of hundred data scientists in India.

The regional teams matter because of how quickly they can work with acquired companies and turn conventional software into AI products. Korawad also argued that Thai teams work effectively across cultures, a capability that becomes operationally important when integrating companies in different countries.

This separates the market that generates revenue from the base that generates execution capacity. A Thai company does not need to earn most of its revenue at home for its Thai and Southeast Asian teams to remain central to its global competitiveness.

Ten years of groundwork before revenue took off

Korawad said on stage that it took Amity about ten years, from 2012 to 2022, to reach $10 million in revenue. He then described roughly two more years, from 2022 to 2024, to reach $100 million.

The year of that $100 million milestone appears in two versions. In Amity's Series D announcement of March 25, 2026, Korawad is quoted saying that annualized revenue, meaning revenue calculated at the current yearly rate, broke $100 million for the first time in 2025, expanding more than 10x since 2022. The session version puts the milestone in 2024, while the company's announcement puts it in 2025.

For 2026, he said Amity was closing in on $200 million in revenue. That matches the target in the announcement, which says the company is aiming to break $200 million in annualized revenue in 2026. The final result remains for the company to disclose. All revenue figures from the session are Korawad's own account.

His lesson was perseverance. Had the company stopped during its first decade, it would not have reached the point where the rollup strategy could begin compounding its earlier investment in research, teams and operating experience.

Korawad also pointed to open source and open weight models as an increasingly practical choice for enterprises. Open weight models make their trained parameters available for organizations to deploy and adapt. He argued that enterprises can train such models on their own data, retain more control over the AI stack and reduce token costs instead of relying exclusively on large frontier model providers.

This brings the discussion back to Vertical AI. Thai founders do not necessarily need the largest model if they can choose a valuable niche, secure differentiated data and build for a global market from the outset. Thai enterprises face a related decision: which data should remain under their control, which tasks warrant specialized models and what operating costs will accumulate over time?

Amity's story is therefore more useful as an operating thesis than as a fundraising milestone. For founders, global ambition has to be matched by a mechanism for acquiring distribution and data. For enterprises, greater ownership of the technology may begin with choosing the model that fits the work, rather than the model with the largest parameter count.

Source: "$160M Raised: Building a Global AI Player from Thailand," a session by Korawad Chearavanont, Founder and Executive Chairman of Amity, at Techsauce Global Summit 2026 on August 28, 2026.

Thailand is hosting the Annual Meetings from 12 to 18 October 2026, marking its first time as host in 35 years, since 1991. The forum will be held on Wednesday, 14 October 2026, from 10:00 to 12:00 at C asean, ThaiBev Quarter, Bangkok, and will convene more than 200 attendees, including senior executives from the financial, investment and business sectors, as well as bank clients, business owners, and local and international partners.

The event addresses critical global transitions: the reallocation of long-term global capital amid geopolitical and climate pressures, the rise of AI redefining competition among countries and organizations, and the shortening climate deadline. The forum seeks to answer whether Thailand and ASEAN will be "players" or merely "spectators" in this new global economic game.

Event Highlights:

• Special Keynote: "Reimagining Trade for the Future: Unlocking Growth in a Changing Global Economy" by Mrs. Suphajee Suthumpun, Deputy Prime Minister and Minister of Commerce of Thailand, discussing the transformation of Thai trade amid the rewriting of global trade rules.
• Global Panel Discussion: "Powering The Future Economy – Growth, Technology, and Sustainability" featuring:
  • Mr. Sarvesh Suri, Regional Vice President for Asia and the Pacific, International Finance Corporation (IFC), World Bank Group
  • Mr. Jan Metzger, Global Head of Coverage Banking, Corporate & Investment Banking, Standard Chartered
  • Mr. Azar Koulibaly, Regional Vice President and Associate General Counsel, Corporate, External and Legal Affairs, Microsoft Asia
  • Mr. Anoop Sagoo, CEO, Southeast Asia, Accenture
  • Moderated by Dr. Wit Sittivaekin

Ms. Kattiya Indaravijaya, Chief Executive Officer, KASIKORNBANK, stated, "Thailand hosting the 2026 IMF-World Bank Group Annual Meetings is a crucial opportunity as global financial and economic leaders gather in Bangkok. KBank aims to use this platform to connect Thai businesses directly with those shaping global capital, technology, and development. We believe growth in the next decade requires finance, technology, and people moving in the same direction. The bank is ready to be the bridge for our customers and Thailand to transition into a future economy that grows sustainably and inclusively."

Ms. Oranuch Lerdsuwankij, Co-founder and CEO of Techsauce, said, "The world is facing three simultaneous transitions: capital reallocation, the advent of AI, and the climate deadline. No single country, organization, or institution can handle this alone. As a driver of Thailand's tech ecosystem, Techsauce has partnered with KBank to design this forum as a highly relevant and actionable dialogue. This will provide Thai business leaders with a clear picture of how to position Thailand and ASEAN in this defining decade."

The "KBank x Techsauce: Empower The Future Economy" event is strictly by invitation only.