Agentic AI: High awareness but low adoption
Next-generation technologies such as agent AI, advanced automation, and robotics are expected to define the next stage of adoption, but readiness remains low. Only 19% of companies say they are completely or very ready to implement agent AI, and just 24% say they have heard of agent AI.
Only 4% of people familiar with this technology have fully adopted it, but those who have adopted it report faster decision-making (67%), increased productivity (63%), and increased operational scalability (54%). Startups are ahead of the curve, with 47% reporting they are ready, compared to 24% of large enterprises and 17% of small businesses.
Skills shortages remain the most commonly cited barrier to deeper adoption, specifically for next-generation technologies, cited by 56% of companies and 51% overall, up from 47% last year.
While software developers, data engineers, and solutions architects are the most in-demand technical roles, the report says the greater new need is human judgment. 61% of companies recognize that the ability to interpret, validate, and challenge AI output is a critical skill for the future, but only 24% are confident that their current training is sufficient.
Batsun said domain experts, employees with deep business knowledge, rather than technical experts, are best suited to spot errors in AI output, urging companies to build clear information. “Human relations” Move checkpoint and escalation processes to AI-driven workflows.
Governance struggles to keep pace
The use of AI is increasingly starting outside the IT department, with 71% of companies saying use cases are currently being driven by business units or non-technical staff, and 70% saying non-technical employees frequently use AI tools.
However, governance has not caught up. Only 19% of organizations have comprehensive and mandatory policies covering non-technical AI uses, 47% have informal or inconsistent policies, and 34% have no policies at all. The report says this gap is leading to the risk of unauthorized use of AI. “Shadow AI” Used by staff to feed company data into their personal AI subscriptions.
The report notes that several organizations in Thailand are already extending AI deeper. Kasikorn Business Technology Group Modernized traditional COBOL-based banking systems using agent AI tools, reducing engineering effort by 2-3x. Consumer goods group Osospa We built OsotSphere, a Thai language assistant that currently handles approximately 1,450 internal queries per month across finance, legal, and marketing. and chiang mai university’s Generative AI platform serves over 52,000 staff and students, with 86% of students already using AI in their learning.
Infrastructure and competitiveness
Launched last year, the AWS Region in Thailand currently offers around 120 services, covering more than 80% of local customer demand and helping enterprises increase their cloud adoption rate from 57% to 64%.
Batsun said enterprise customers are increasingly migrating workloads from Singapore to Thailand, where baseline infrastructure costs are about 10-15% cheaper.
Overall, 54% of businesses rate Thailand as a competitive hub for AI and innovation, citing strong digital infrastructure and a large consumer market, while 56% cite a lack of digital talent and 51% cite limited access to funding and venture capital as a constraint to competitiveness.
Regulatory uncertainty was also cited, with 41% citing legal uncertainty and 44% citing regulatory complexity as a barrier, while 47% said clearer regulations are important for AI adoption.
The report calls for a harmonized national AI roadmap to give businesses long-term certainty around regulation, infrastructure and investment. Expanded employee training to focus on applied judgment rather than basic digital literacy. A stronger AI governance framework with clear escalation routes. and a regulatory environment that fosters innovation in line with international standards.
“The challenge for Thailand is no longer simply to increase adoption rates.” The report concludes: “But it allows companies to move from experimentation to transformation.”
Without this change, they warn, the benefits of the next wave of AI will be concentrated among a few advanced companies, leaving the rest of the economy with shallow, incremental gains.
