Thailand must turn artificial intelligence (AI) adoption into measurable business returns by redesigning operations, strengthening human oversight and upgrading workforce skills, according to speakers at the KBTG Techtopia seminar.
Failing to do so risks widening the gap between large companies and smaller firms, they noted.
As AI moves from answering questions to executing tasks, businesses face pressure to control costs, set clear limits on agents’ authority and retain human approval for critical decisions, particularly in financial services.
The technology executives urged companies to focus investment on business priorities, equipping employees for higher-value work, while the government called for broader access to digital skills and infrastructure to strengthen Thailand’s competitiveness.
QUANTIFY THE OUTCOME
“AI agents are expanding technology’s role from conversational assistance to planning, reasoning, design and execution,” said Voranuch Dejakaisaya, executive chairman of Kasikorn Business-Technology Group (KBTG).
KBTG has developed a multi-agent platform called AthenaMind and 90% of employees across KBTG actively use AI in their work.
AI agents work alongside engineers on requirements, gathering, design, coding and testing, with KBTG targeting a roughly 15% reduction in end-to-end software development lifecycle costs by 2028.
The target links AI deployment to a measurable business outcome across the development process, Ms Voranuch said.
The company has deployed more than 80 AI projects, covering applications such as knowledge management, document processing and voice processing.
However, Ms Voranuch warned that adoption must be accompanied by controls appropriate to the business risk. Core banking and payment systems cannot operate with fully unmonitored AI autonomy.
“Some statistics estimate only one in five companies have mature governance for autonomous AI agents,” she said.
There needs to be clean, reliable data verification across multiple sources as well as measures to reduce training data bias, while retaining human authority over final decisions, said Ms Voranuch.
As AI automates code generation, traditional programming roles are changing. Employees must learn to use AI effectively to create business value, alongside developing creativity, empathy and leadership, she noted.
Under its “AI First, Human First” approach, KBTG uses internal governance tools including SaiJai, which evaluates whether AI operations remain within designated parameters, and AEGIS, which enforces security policies and controls.
The company’s AI Playground provides a monitored environment for employees to test tools before wider deployment.
Ms Voranuch outlined a tiered approach to oversight: low-risk tasks can be automated, medium-risk activities require a human reviewer, and critical decisions remain subject to full human discretion.
Each agent should have clearly defined responsibilities and limits on its authority. Organisations also need an independent external kill switch capable of halting an agent when necessary.
Safety testing must examine more than whether an AI response is correct. It should establish how far a failure could spread before controls contain it, supported by the ability to detect problems, stop operations and recover quickly.
Ms Voranuch said the golden rule of operational resilience when managing autonomous AI agents is that when an agent makes a mistake or behaves unpredictably, organisations must detect the anomaly immediately, pull the plug cleanly, and restore stability fast.
REDESIGNING BUSINESS
Ruangroj Poonpol, co-founder of House of Wisdom, said companies should distinguish business transformation from task automation, with investment focused on customer needs, operating costs and organisational change.
House of Wisdom is a private, invitation-based circle of more than 300 leaders in Bangkok.
“Are we transforming the business, or are we just automating the tasks?” Mr Ruangroj asked.
Businesses should select one to three priority areas, integrating AI deeply before expanding, rather than pursuing fragmented experiments.
“Select bets that are big enough to matter, but small enough to win,” he said.
A shared technology platform would allow capabilities to be reused across the enterprise, while model selection should reflect both task requirements and running costs.
“It’s no longer just ‘put the right man into the right job’. More important is to ‘put the right AI into the right job at the right token’,” said Mr Ruangroj.
Empathy, care, mentoring and an open-ended strategy should remain human-led, while bounded tasks offer greater automation potential, he noted.
Companies could retain entrepreneurial talent by providing capital and shared ownership to develop businesses internally. However, Mr Ruangroj cautioned against treating workforce reductions as the defining achievement of AI adoption.
“Don’t let your legacy be telling your grandchildren that, ‘Grandpa brought in AI and laid off 1,000 people.’ Never let that be your legacy,” he said.
Transformation must be carried out by and for humanity, said Mr Ruangroj.
CLOSING THE DIVIDE
“We must leap past basic digital adoption to achieve true digital transformation,” said Deputy Prime Minister Suphajee Suthumpun at the event.
Thailand must embed AI and data analytics into business operations and upgrade workforce skills to prevent the gap between large companies and SMEs from widening.
She warned that an uneven access to technology could deepen the uneven economic recovery as geopolitical tensions, export concentration and a shrinking workforce strain competitiveness.
The government is connecting data from 23 agencies, expanding practical digital training through the DBD Academy and streamlining business permits. It also aims to link SMEs with larger companies across supply chains.
The country needs to transition from a standard tourism model to a visitor economy tailored to medical, educational, and long-stay visitors, while attracting global tech talent, said Mrs Suphajee.
The transformation also requires stronger digital infrastructure, cybersecurity and governance, alongside clean energy for data centres and efforts to attract technology talent, she noted.
“The economy must be driven by innovation, businesses must compete globally with fairness, and Thais must possess skills ready for the future,” said Mrs Suphajee.