KBTG has developed its multi-agent platform, Athena Mind, as part of this transition towards a digital workforce. More than 90% of employees at KBTG and KASIKORNBANK can now apply AI to their work, while KBTG has expanded AI deployment to more than 80 projects.<\/p>\n\n\n\n
AI has also been introduced into the software development process through AI SDLC, with KBTG targeting a 15% reduction in end-to-end software development costs by 2028.<\/p>\n\n\n\n
As AI becomes capable of taking more action independently, however, defining its authority becomes increasingly important.<\/p>\n\n\n\n
KBTG has adopted a \u201cHuman First x AI First\u201d approach, maintaining humans in the loop for critical systems. Its governance framework includes a Control Tower to define the roles and authority of AI agents, an AI Kill Switch that allows systems to be stopped externally, and Red Team testing designed to examine their limits.<\/p>\n\n\n\n
The objective is not to accumulate the largest number of AI projects or use cases, but to generate business value while keeping humans in control.<\/p>\n\n\n\n
\u201cWe have to be the master of AI, rather than allowing AI to become our master,\u201d Voranut said.<\/p>\n\n\n\n
Thailand must move from AI user to AI creator<\/h3>\n\n\n\n
At the national level, the challenge extends beyond encouraging more people and organisations to use AI. Thailand must also develop the capability to create technology, innovation, businesses and talent for the new economy.<\/p>\n\n\n\n Prof. Dr.Yodchanan Wongsawat, Deputy Prime Minister and Minister of Higher Education, Science, Research and Innovation<\/strong>, outlined a \u201cTech for Thailand\u201d direction in which research and development and innovation are used both to improve the productivity of existing industries and to create higher-value industries as new economic engines.<\/p>\n\n\n\n That requires research, technology, people, capital and businesses to be connected, alongside the upskilling and reskilling of workers for higher-value industries. Research must ultimately generate innovation and revenue that can be reinvested into further R&D.<\/p>\n\n\n\n Thailand has already seen a sharp increase in organisational AI adoption. According to the AI Readiness Survey 2026 presented by Dr.Chai Wutiwiwatchai, Executive Director of the National Electronics and Computer Technology Center (NECTEC)<\/strong>, the proportion of organisations using AI rose from 17% in 2024 to 58% in 2026.<\/p>\n\n\n\n Technology capability, however, remains a weakness. Thailand therefore needs to move from being a \u201csuper user\u201d towards developing AI technologies and businesses of its own. Chai proposed large national programmes that bring multiple organisations together around clearly defined challenges, allowing research to move beyond proof of concept and scale into businesses and regional markets.<\/p>\n\n\n\n Dr.Krithpaka Boonfueng, Executive Director of the National Innovation Agency (NIA)<\/strong>, also sees potential for Thailand to turn AI into solutions and businesses, but this requires stronger connections between research, technology, startups and markets, as well as opportunities for Thai innovations to be tested and grow.<\/p>\n\n\n\n \u201cIf we do not give Thai people and Thai innovations an opportunity, we will remain users forever,\u201d she said.<\/p>\n\n\n\n For businesses, the challenge is also changing. The question is no longer whether an organisation has AI, but whether AI can produce measurable business results.<\/p>\n\n\n\n Krating Ruangroj Poonpol, co-founder of House of Wisdom<\/strong>, questioned why many organisations have invested in AI, built numerous use cases and introduced copilots, agents, chatbots and AI labs, yet still see little or no impact on their profit and loss.<\/p>\n\n\n\n He described this situation as \u201cAI Theater\u201d \u2014 organisations ap<\/p>\n\n\n\n pearing highly active in AI without translating that activity into meaningful business impact.<\/p>\n\n\n\n His answer is to \u201cdo less\u201d, rather than continually adding more projects.<\/p>\n\n\n\n Instead of operating a use-case factory, organisations should select one to three important domains and transform them deeply enough to generate measurable results.<\/p>\n\n\n\n Simply adding AI to existing companies and existing processes risks doing little more than \u201cautomating the past.\u201d AI-native transformation requires organisations to reconsider how they would design their businesses if they were building them today with the capabilities AI now provides.<\/p>\n\n\n\n Anothai Wettayakorn<\/strong>, Managing Director and Technology Leader of IBM Thailand<\/strong>, similarly argued that the difference between organisations generating value from AI and those still stuck in pilots does not necessarily lie in budget, workforce or model size.<\/p>\n\n\n\n Leadership mindset matters. An AI-first enterprise must move from experimentation towards integration, connecting AI with applications, data, workflows, automation, security, governance, people and skills before progressing towards transformation of business processes and business models.<\/p>\n\n\n\n The emergence of agentic AI introduces a new organisational challenge. When AI can receive goals, break them into tasks and continue working towards an outcome, building the first agent is no longer the difficult part.<\/p>\n\n\n\n Vatsun Thirapatarapong, country manager, AWS Thailand<\/strong>, said the larger challenge is enabling multiple agents to work together \u2014 connecting information across systems, coordinating tasks, defining access rights and providing mechanisms to investigate failures.<\/p>\n\n\n\n Organisations will increasingly need to manage AI agents in much the same way they manage employees, with policies and identities defining which data and systems each agent is authorised to access and records showing what each agent has done.<\/p>\n\n\n\n The central question is therefore shifting from whether an organisation can build an AI agent to whether it can manage multiple agents as a functioning workforce.<\/p>\n\n\n\n Quality assurance must evolve accordingly. Thadpong Pongthawornkamol, Managing Director of KBTG<\/strong>, said AI requires a shift from one-time certification towards lifetime evaluation, in which systems are continuously monitored, tested and assessed throughout their lifecycle.<\/p>\n\n\n\n KBTG is also using AI to evaluate AI through Sai JAI, an LLM judge, while retaining human oversight. The traditional tester consequently evolves towards the role of a \u201cQuality Guardian\u201d, with humans continuing to hold responsibility for quality decisions.<\/p>\n\n\n\n Moving AI from pilot projects into enterprise-wide deployment requires more than selecting the right model. Organisations also need reliable data, scalable infrastructure, governance, security and people capable of operating the systems.<\/p>\n\n\n\n The foundation for scaling AI is knowing where data resides, which systems are being used and how they are connected, while establishing from the outset how AI will generate financial impact, productivity gains, adoption and trust.<\/p>\n\n\n\n Security becomes even more critical as organisations move towards agentic AI. People with legitimate access to information can themselves become attack vectors, making awareness, Zero Trust and cybersecurity essential components of AI deployment.<\/p>\n\n\n\n In this environment, cybersecurity effectively becomes a \u201clicence to scale\u201d \u2014 the foundation that allows organisations to expand AI while maintaining security and trust.<\/p>\n\n\n\n The Human of Tomorrow discussion also extends beyond workplaces and economies to the cities in which people will live.<\/p>\n\n\n\n Karndee Leopairote, an MP and specialist in technology, foresight, digital economy, innovation and policy design<\/strong>, argued that smart cities should not begin with technology.<\/p>\n\n\n\n They should begin with an understanding of how people will live in the future.<\/p>\n\n\n\n Families are becoming smaller, people are living longer, birth rates are declining and patterns of work are changing. Cities must simultaneously cope with heat, flooding and ageing populations.<\/p>\n\n\n\n Urban development therefore needs to incorporate nature-based planning, climate data and local conditions, alongside concepts such as the 15-minute city, where hospitals, schools and essential services can be reached close to home.<\/p>\n\n\n\n Technology alone cannot solve urban problems. Civic co-creation and open data are equally important, allowing citizens to access information, scrutinise decisions, offer suggestions and participate in designing their cities.<\/p>\n\n\n\n Technology, in this view, is a tool for building cities around human lives rather than an objective in itself.<\/p>\n\n\n\n Ultimately, the Human of Tomorrow conversation returns to the individual. If machines can increasingly perform tasks once carried out by humans, what capabilities should people protect and develop?<\/p>\n\n\n\n Pat Pataranutaporn, a technologist at MIT Media Lab and founder and director of the Cyborg Psychology Research group<\/strong>, suggested that one of AI\u2019s risks may not be the familiar image of machines destroying humanity.<\/p>\n\n\n\n A more immediate danger is that technology could erode aspects of being human if people gradually think, learn and make decisions less for themselves.<\/p>\n\n\n\n Cyborg Psychology therefore examines how AI changes human psychology, learning, questioning, decision-making and relationships.<\/p>\n\n\n\n The focus shifts from making AI more intelligent towards intelligence augmentation \u2014 using technology to make humans more capable \u2014 and ultimately towards AI for human flourishing.<\/p>\n\n\n\n Thana Thienachariya, Director and Chairman of Bluebik Group, suggested that people may not need to compete with AI by trying to know more about technology than machines do.<\/p>\n\n\n\n Instead, they can accumulate \u201cdots\u201d through diverse skills, experiences and relationships \u2014 a process of skill stacking that allows those dots to be connected into ideas, self-understanding and problem-solving capabilities.<\/p>\n\n\n\n Curiosity, discipline and empathy consequently become increasingly valuable human qualities.<\/p>\n\n\n\n Technology should not determine what people are curious about. Skills still require the discipline to practise. And human value extends beyond capability to relationships and the ability to understand others.<\/p>\n\n\n\n Writer Sarawut Hengsawad, known as Roundfinger<\/strong>, raised another question: if AI saves people time, what happens to that time?<\/p>\n\n\n\n If every hour saved through technology is simply filled with more work, humans remain trapped in an endless cycle of optimisation.<\/p>\n\n\n\n AI, from this perspective, should not only increase productivity. It should also give time back to people \u2014 time to live and reconnect with those around them.<\/p>\n\n\n\n As AI becomes more powerful, individuals must also become more aware of how technology can influence their perceptions and emotions.<\/p>\n\n\n\n AI-generated images, voices and identities are becoming increasingly realistic, meaning that seeing or hearing something may no longer be enough to establish whether it is genuine.<\/p>\n\n\n\n Trynh Phoraksa, Criminology Lecturer, Faculty of Social Sciences and Humanities, Department of Social Sciences at Mahidol University<\/strong><\/strong>, noted that intelligence or education does not necessarily protect people from scams.<\/p>\n\n\n\n \u201cScammers don\u2019t play with your IQ. They play with your EQ,\u201d he said.<\/p>\n\n\n\n Fear, greed or emotional attachment can temporarily reduce a person\u2019s ability to analyse information and weigh evidence, creating opportunities for social engineering. AI and deepfakes can make such manipulation even more powerful.<\/p>\n\n\n\n The Human of Tomorrow therefore needs more than the ability to use AI. People also need to understand what technology can do to their own perceptions and emotions \u2014 and retain the ability to stop and think before making decisions.<\/p>\n\n\n\n At the system level, trust can no longer be assumed. Jessica Rosenworcel, former Chairwoman of the FCC<\/strong>, said the early internet was built on the assumption that network nodes could be trusted. That assumption is no longer sufficient in a world where machines can increasingly operate without humans directly involved.<\/p>\n\n\n\n Trust needs to combine credibility, reliability and safety, while giving humans enough information to examine systems and decide which technologies should become part of their lives and work.<\/p>\n\n\n\n Trust, therefore, needs to be designed into technology from the beginning rather than added afterwards.<\/p>\n\n\n\n KBTG Techtopia extended these ideas beyond its three conference stages through workshops, Community Circle sessions and technology showcases.<\/p>\n\n\n\n The workshops ranged from application development with Google AI Studio and Google Stitch to AI-assisted legacy code improvement, AI agent skills with IBM Bob, security and compliance applications involving Splunk, and AI-driven software development.<\/p>\n\n\n\n Other sessions addressed the design of trustworthy AI products, psychological safety, leadership, listening and human relationships.<\/p>\n\n\n\n The Community Circle brought the discussion closer to everyday life, exploring careers, personal strategy, behavioural change, burnout, AI adoption, entrepreneurship, relationships and collaboration between people of different generations and perspectives.<\/p>\n\n\n\n Technology showcases from KBTG and its partners provided another link between ideas and practical applications.<\/p>\n\n\n\n One example was KBTG\u2019s AI SDLC, which introduces AI throughout the software development lifecycle rather than limiting it to coding.<\/p>\n\n\n\n The initiative began as AI Augmented SDLC in late 2024 and early 2025, with the objective of accelerating the entire software development process while maintaining quality and human involvement at critical stages.<\/p>\n\n\n\n Different agents assist with requirements, functional and technical specifications, task classification, coding, code review and test-case generation. Tasks that AI cannot yet handle are passed to humans, supported by AI coding assistants.<\/p>\n\n\n\n The impact on speed varies by task. Some coding work that previously took around two days can be reduced to about an hour, while design work requiring more human thinking and joint decision-making has seen improvements of only around 10%.<\/p>\n\n\n\n The principle remains human-in-the-loop. AI takes on routine work and helps identify issues people may overlook, while humans review, decide and concentrate on more difficult and higher-value work. The intention is not to replace people, but effectively to give every person additional digital support.<\/p>\n\n\n\n Taken together, the discussions at KBTG Techtopia 2026 point towards changes that need to happen simultaneously at several levels.<\/p>\n\n\n\n Thailand needs to move from being primarily an AI user towards creating its own technologies and innovations, connecting R&D, talent, capital, startups and businesses.<\/p>\n\n\n\n Cities need to use technology to address how people actually live while giving citizens a role in shaping their communities.<\/p>\n\n\n\n Businesses need to move beyond AI pilots and use-case accumulation towards measurable business impact, supported by data, platforms, security and governance.<\/p>\n\n\n\n The rise of agentic AI and digital workforces means humans themselves will increasingly shift from performing every step of a process towards setting objectives, supervising systems, making decisions and remaining accountable for outcomes.<\/p>\n\n\n\n For individuals, being the Human of Tomorrow does not mean becoming the person who is best at using AI.<\/p>\n\n\n\n It means continuing to think independently, ask questions and verify what machines produce. It means preserving curiosity, discipline, empathy, experience and relationships. It means understanding how technology can influence emotions and judgement, and knowing when technology should enhance human capability \u2014 and when it should be put aside to make room for life.<\/p>\n\n\n\n As AI evolves from a tool into systems capable of reasoning, planning and taking action, humans do not need to become more like AI or race against its capabilities.<\/p>\n\n\n\n The more important question is what humans choose to build with those capabilities \u2014 while retaining the authority to decide where technology should take them.<\/p>\n\n\n\n The ultimate goal is not to allow AI to become humanity\u2019s master, but to ensure that humans remain the \u201cmaster of AI\u201d, using technology to create value for business, the economy, society and human life.<\/p>\n\n\n\n![\"\u0e28.\u0e14\u0e23.\u0e22\u0e28\u0e0a\u0e19\u0e31\u0e19](\"https:\/\/www.thestorythailand.com\/wp-content\/uploads\/2026\/09\/1790082189_831740-thestorythailand.avif\")
The age of AI experimentation is ending<\/h3>\n\n\n\n
AI agents are becoming part of the workforce<\/h3>\n\n\n\n
![\"KBTG](\"https:\/\/www.thestorythailand.com\/wp-content\/uploads\/2026\/09\/1790081522_651153-thestorythailand.avif\")
<\/figure>\n\n\n\nScaling AI requires more than powerful models<\/h3>\n\n\n\n
Smart cities should begin with people, not technology<\/h3>\n\n\n\n
As AI becomes smarter, humans cannot afford to think less<\/h3>\n\n\n\n
Trust must be designed into technology<\/h3>\n\n\n\n
![\"Jessica](\"https:\/\/www.thestorythailand.com\/wp-content\/uploads\/2026\/09\/1790081600_371166-thestorythailand.avif\")
Putting Human of Tomorrow into practice<\/h3>\n\n\n\n
Humans must remain the ones setting the direction<\/h3>\n\n\n\n