Experts at Boston Consulting Group (BCG) said that for corporations to succeed in the age of artificial intelligence (AI; Artificial Intelligence), investment responsibility should be distributed across the organization rather than concentrated in a single chief executive officer (CEO). It means the division of roles should place ultimate responsibility for setting the vision and delivering results on the CEO, with the management team executing and implementing that vision, and the board overseeing and verifying the process. They also advised shortening the AI decision-making cycle and making the CEO's accountability clearer.
The following is a Q&A with BCG managing director and senior partners Alpesh Shah and Jeff Walters.
─ Although CEOs and boards generally agree on the importance of AI, why does such a large gap emerge at the execution stage?
"CEOs and boards agree on the importance of AI, but they have different expectations about the value AI can actually create and the pace of execution. Some CEOs believe boards overestimate AI's potential or do not fully understand its impact on growth strategy. Boards, in turn, often feel that corporations are not moving fast enough in their AI transformations.
These perception and expectation gaps can lead to unrealistic expectations about what AI can actually deliver, widen the gap during execution, and create tension between rapid AI adoption and appropriate oversight. To fix this, CEOs and boards need to talk about AI in the same language."
─ From what perspective should CEOs manage AI return on investment (ROI; Return on Investment), and how should boards evaluate it?
"CEOs should embed AI within corporate strategy. Rather than running hundreds of pilot projects, it is important to focus on core functions with high value-creation impact. BCG categorizes how to maximize AI value into 'deploy' to boost productivity, 'reshape' to change workflows, and 'invent' to create new products and business models.
Boards should also evaluate AI using the same standards as capital investments. The question is not simply whether AI was adopted, but whether AI actually reduced unnecessary work, streamlined organizational layers, improved the customer journey, and contributed to gaining market share."
─ How are the standards for a 'good board' changing in the AI era?
"AI literacy has now become a basic requirement for board members. A good board goes beyond overseeing risks and financial performance to act as a thought partner that validates the CEO's key AI-related decisions and weighs the pros and cons of major choices together. And as AI becomes more deeply embedded across operations and regulation tightens, the board's role in managing AI-related risks—such as cybersecurity, data governance, and regulatory compliance—is becoming more important."
─ As AI shifts from a technology project to a business transformation mandate, the CEO's role is changing too. What leadership capability matters most for CEOs in the AI era?
"CEOs must be able to show how they will create value using AI. Rather than simple adoption, they should focus on redesigning work and businesses and creating new AI-based products and business models. Given the high uncertainty around AI, CEOs must manage long-term judgment and short-term performance at the same time."
─ How should the division of roles among the CEO, the board, and the management team be designed in the AI era?
"For AI to succeed, accountability should be distributed across the organization rather than concentrated in one CEO. It is desirable for the CEO to be accountable for vision and results, the management team to handle execution and deployment, and the board to perform oversight and verification.
What matters is that the CEO does not necessarily have to be a technical expert. It is also possible to have a co-CEO model with a technical background, or to have a strong chief technology officer (CTO) or chief AI officer (CAIO) support the CEO's AI strategy and transformation journey."
─ What leadership blind spots do corporations most often miss during AI transformation?
"The most common mistake is confusing AI adoption itself with transformation. Simply introducing AI tools is not enough. Decision-making authority and incentive structures must also change.
Many also see AI only as a technology investment and fail to consider the data foundation, operating model, and workforce capability building. When early results appear, leaders often limit investment before the transformation is complete. Lastly, treating AI readiness as a simple HR task is another classic leadership blind spot."
─ What are the basic elements of AI governance that corporations should put in place first?
"First, establish a value accountability system. Create a formal transformation organization that spans finance, technology, operations, HR, and risk, and reflect KPIs (key performance indicators) tied to profit and loss and accountability in incentives. Next is capital discipline. You need a decision-making system that checks whether operational improvements are actually translating into financial performance before deploying additional investment. What underpins these two is the board's practical understanding of AI. For Korean corporations, with the Basic AI Act coming into force, it is important to build a governance framework that internalizes regulation from the outset rather than applying it retroactively."
─ What should Korean CEOs and boards prioritize right now, and what decision-making practices should change to drive a successful AI transformation?
"What matters is not the speed of investment but the structure that produces results. The priority now is to industrialize AI within actual workflows. Rather than multiplying scattered pilot projects, you should focus on one to three core functions that can continuously affect profit and loss.
Given the Korean government's AI direct investment plans this year, two changes are essential. First, shorten the AI decision-making cycle. Korea's consensus-centered governance model has strengths for long-term agendas but can be a constraint on AI's pace. Second, make the CEO's accountability clearer. Set metrics tied to profit and loss and review them regularly. Ownership without accountability is just a title."