“Only around 7% of firms currently use AI.” – US Census Bureau, May, 2025
“Some leaders will pursue bottom-line gains strategically; others won’t — and will crash into the limits of quick fixes.” – Forrester, October, 2024
“The gap between what CEOs need to know and what they actually understand is massive.” – Glenn Hopper, Business Insider, June, 2025
“AI adoption is still in its early innings.” – Andrew Ng, Bloomberg, May, 2025
These four quotes reveal a paradox: while AI progress accelerates technologically, adoption within organizations remains slow and uneven.
To the average person who has more than a passing interest in AI, the hype about the near-arrival of AGI (artificial general intelligence) can feel overwhelming. But to the average employee or business leader in a corporation, even large ones with substantial IT budgets, the pace of AI adoption can feel underwhelming.
There are many reasons for slow organizational adoption: shortage of AI skills, perceived risk of proprietary and personal data leakage, anxiety over job loss, or lack of clarity over ROI.
A significant reason, which underpins the above reasons, is a lack of leadership. If leadership does not tie AI initiatives to line-item results, or set a compelling vision for a better way of working, culture will devour your AI strategy for breakfast, lunch and dinner.
The SCBX Case
It was August, 2024 and Dr. Arak Sutivong, Deputy CEO of leading Thai financial technology group, SCB X PCL, was presenting at Techsauce in Bangkok, Thailand.
Everyone was talking about AI, but Sutivong, who will take over as SCBX’s next CEO from January, 2027, was one of the few talking about actual AI adoption.
While many corporate leaders aspire to be AI First companies, Sutivong shared how SCBX is AI First, displaying four specific targets:
- 75% of revenue from AI by 2027
- 100% of SCBX Group employees attain basic AI literacy by 2025
- 15% of Group employees have advanced AI capabilities built and leveraged through data and R&D
- Responsible AI policy and system developed and deployed
Sutivong explained that SCBX, an investment holding company and a parent company of 118-year old Siam Commercial Bank, which serves 200 million customers across Asia, had set these group metrics back in 2023, when leadership realized AI was a very natural progression from being a digital organization.
Digital transformation has been their focus through the 20-teens in order to provide the kind of customized and timely service customers wanted and expected. But once AI entered the picture, they understood that digital was not just a channel to the customer – it was a way to create value. “Once we realized that,” said Sutivong, “the group CEO said AI had to be the tone and direction.”
Anchor to Value and Vision: SCBX’s AI Impact on Revenue
Impactful leadership is about setting goals in blue oceans, when you’re not sure how you’re going to get there. It’s bold to say 75% of SCBX’s P&L will be enabled by AI via enhanced revenue generation and cost cutting, but how exactly do you determine AI’s impact?
“It’s in the DNA of our organization to set a very clear, compelling, and seemingly stretch vision and target to send the signal to the organization that this is real and important,” explained Sutivong. “But (measuring progress) is still a journey, a work in progress.”
Sutivong said that AI adoption is being driven jointly by an AI Center of Excellence (AICOE), which provides the technical expertise and guidance on AI tools and processes, and the Group CFO, who oversees implementation and is “the referee” who tracks progress towards plan.
The Group CFO engages all group functions and businesses in discussion about critical workflows, particularly those which impact revenue enablement, cost reduction or process improvement, to determine what the potential AI penetration could be.
To do this, they undertake a “painstaking” process for each business, such as the mortgage business, breaking it down “from origination all the way to collection”. This involves identifying how many steps one takes in a given process, and how many of those steps AI can have an impact.
For example, the Group CFO would ask in the mortgage business, what steps do we take from origination to collection? How many steps are there? Which steps can be done more efficiently or effectively by an AI tool or technique? Would it be in the initial evaluation phase of the customer and the loan, the data processing on the online platform, the submission of the application, the preparation of the loan proposal, the underwriting report, the compliance checks, etc. etc.?
After some re-engineering is applied, the Group CFO would then assess whether a “sufficient amount of AI penetration” has been applied. To quantify AI’s enablement of the P&L, they divide the number of steps taken over by an AI tool or technique by the total number of steps in the process. For example, if say, 6 out of 15 steps are enabled by AI, then that 40% would be rolled into the aggregate SCBX Group AI penetration total.
Sutivong said that all up they are currently around 35% AI enablement, but still optimistic of getting to 75% by the end of 2027. For now, it’s important that all employees understand what they mean by AI First.
“It’s not completely precise, but it definitely helps everyone understand how we count,” said Sutivong. “And everyone understands that we have this third party, Group Finance, refereeing this whole thing.”
Empower the Workforce: Educate
One can say that the quick-and-dirty playbook for AI Adoption in general is two-fold: the top-down leadership communication and enablement, and the bottom-up learning, experimentation and sharing.
SCBX very quickly put together a curriculum and learning process for foundational, intermediate and advanced levels through collaboration with technology and knowledge partners like Microsoft and BCG.
The AI Foundation course was developed by the SCBX Talent Office, and rolled out in June of 2024. This mandatory course is an online program delivered in bite-sized chunks that ultimately runs about 90-minutes long.
The goal of the foundational course is to help all employees understand that AI tools should be as second nature to use as Microsoft Excel or PowerPoint. As Sutivong said, “they need to know what the tools are, what they can do, as well as what they cannot do.”
To enable more hands-on application, the SCBX Talent Office produced a secondary level of education, called AI Intermediate program. At this level, employees who have their own AI use case ideas, can apply for a development program where a cohort of 60 to 70 people come together, and are divided into about 15 project groups. The groups are coached by AI specialists and domain experts. Over 300 employees have developed their AI acumen in this 3-month program.
At the AI Intermediate level, employees have learned how to improve and leverage their AI projects, such as functional chatbots, for example, that allow employees to get specific information on HR policies and benefits, have insightful discussions on the company’s structured data, or for sales agents to get investment recommendations to share with their customers.
With the AI Foundation course, SCBX essentially hit their target of 100% of SCBX Group employees attaining basic AI literacy by 2025, at the end of 2024.
Empower the Workforce: Share
More importantly, not only are employees learning about how to use AI. They are sharing their achievements.
Regular gatherings are organized to encourage employees to come together to present their ideas or demo their tools. Above and beyond the education and the hands-on application of AI, sharing ideas and showing demos expand every individual’s perception of how broadly or deeply AI can be leveraged, hopefully igniting their imaginations.
They can think big. And SCBX gives the big dreamers a stage. At the end of 2024, SCBX staged their first AI Battle, an internal contest pitting the best AI use cases across the Group.
One use case highlighted at AI Battle is for market conduct compliance, where a salesperson selling financial products has to inform the customer about several specific things: what the interest rate is, and how you can borrow only what you can pay back, for example.
This requirement mandated by the Bank of Thailand is a challenge to audit. Previously, the authorities would do a mystery shopper sampling to see how consistently bank sales people were compliant.
But thanks to inventive employees at SCBX, sales agents can now record their interactions with the customer (with their permission). Then speech-to-text software converts the conversation to text, and an in-house LLM model validates the conversation to see whether the sales person made the necessary market conduct statements.
If there is an issue, the sales agent gets an email, which helps nudge the agent’s behavior.
“AI frees up our internal audit people, gives comfort to the salespeople, and comfort to the Bank of Thailand,” said Sutivong. “Everyone is operating with peace of mind.”
AI Adoption Principles for Enterprises
In every industry in every major economy, the race is on to understand how to leverage artificial intelligence tools and systems into one’s business. To succeed, and thus potentially race ahead of one’s competitors, leaders need to apply principles of AI adoption, emerging best practices observed across high performing AI-adopting firms globally. I suggest five:
- Anchor to Value & Vision: Tie every AI initiative to a line-item result and a compelling vision for a better way of working.
- Empower the Workforce: Invest as much in the change journey and psychological safety as in the algorithms themselves to turn resistance into advocacy.
- Build a Resilient Data Fabric: Treat your data foundation as a critical change management initiative, not just a technical project, to break down silos.
- Prototype Fast, Scale Smart: Embrace rapid prototyping to experiment and learn; then scale only the proven successes with disciplined AI model lifecycle management.
- Govern for Trust & Velocity: Embed responsible, ethical AI governance from the start to build trust and accelerate, not block, deployment.
In this article, I highlighted the ways in which SCBX has anchored AI to the Group’s value and vision, and how SCBX has empowered its workforce. They have taken actions that demonstrate commitment to the other principles, although not highlighted in this article.
With SCBX’s relatively early commitment to being an AI-First organization, they have created the potential for them to ride the AI S-Curve to greater heights.
As Sutivong notes, leaders today face an accelerating influx of AI capabilities – faster than any team can fully digest. In this environment, a healthy dose of “paranoia,” as Intel’s Andy Grove once said, becomes a strategic necessity – not for leaders and employees to fear AI, but to remain vigilant, adaptive and ready to take advantage of its power effectively and responsibly.
The scary thing is that more and more things start to come out, and it’s forming an unstable queue, where there is more coming in than out, and you cannot digest it all fast enough. And deep in your heart, you also know that there are many smaller fintech equivalents who are actually able to ride this curve and take more risk. There is a little bit of paranoia at work here, but in a sense, that’s how we have to be. And we hope that our processes, and our people will be ready whenever the new and latest come out.
ARTICLE FAQS
- Why do many organizations struggle with AI adoption?
Common challenges include a shortage of skilled talent, concerns about data security and privacy, uncertainty about return on investment, resistance to change, and insufficient leadership commitment to linking AI initiatives with measurable business outcomes. - What does it mean for a company to be “AI First”?
An AI First company integrates artificial intelligence into core business processes and strategy, setting clear targets for AI’s contribution to revenue, efficiency, and decision-making, and ensuring that employees have the skills to work effectively with AI tools. - How can businesses measure the impact of AI?
Organizations can map out key workflows, identify where AI can add value, and track the proportion of process steps improved or automated by AI. These percentages can then be aggregated to estimate AI’s contribution to business performance. - How should companies build AI skills across their workforce?
A structured learning approach works best, starting with foundational AI literacy for all employees, followed by intermediate and advanced programs for those with specific project ideas, and supported by opportunities to share use cases and collaborate. - What principles support successful AI adoption?
Key principles include anchoring AI initiatives to business value and vision, empowering the workforce, building a strong and integrated data foundation, prototyping quickly and scaling proven solutions, and governing AI responsibly to build trust. - What is the AI Adoption framework described here?
The framework is a structured approach for enterprises to adopt AI effectively, based on five pillars: 1) Anchor to Value and Vision, 2) Empower the Workforce, 3) Build a Resilient Data Fabric, 4) Prototype Fast, Scale Smart, and 5) Govern for Trust and Velocity.
