How boards are assessing AI risk for strategic advantage

As AI reshapes the landscape and becomes ever more critical to a company’s strategy and competitive position, a top priority for boards is to understand where to focus their attention to help the company capitalize on AI’s value creation opportunities.

The boardroom has moved past the ‘wow factor’ of generative AI to focus on achieving productivity and ROI. Critical to realizing its promised benefits is the retraining of workers to drive behavioral change and encourage employees to use generative AI to free up time to become more productive. More than three quarters of our board survey respondents say that the top benefit of adopting and leveraging generative AI is optimizing their company’s operations – efficiency, productivity and cost savings.

AI has not leveled the playing field as some had predicted. Leading firms are pulling away in almost every industry because they are using the technology differently. Companies and their boards should expand how they look at generative AI beyond how it can help cut costs to consider how it may be used as a force amplifier to expand the company market or deliver services differently.

Data can be a bottleneck to capturing the benefits of generative AI. Boards need to ask questions about the privacy, governance and quality of the data: Where does the company maintain its proprietary data? How much sits in a cloud-based platform?

While only a modest number of directors report that their companies are deriving revenue from generative AI, there are some industries where it has become key to business strategy, driving innovation and efficiency in product design, supply chain optimization, predictive analytics and personalized customer experiences.

Some companies are adopting or exploring emerging forms of AI such as AI agents to automate and re-engineer workflows. In our Board Leadership Center (BLC) director survey, 23 percent of respondents say the adoption of AI agents and agentic AI is a strategic priority for their company for 2025, while 14 percent say it is not a strategic priority but should be. Another 30 percent say their company is actively exploring emerging forms of AI for adoption over the next two to three years.

Risk and governance: AI agents and agentic AI

The BLC report discusses the risks posed by generative AI – inaccurate data and results, cybersecurity, data privacy and security risks, compliance, intellectual property and reputational risks – as well as guardrails and governance policies for its development, deployment and use. Companies deploying the technology should take a hard look at their data quality and governance practices: achieving productivity and efficiency improvements with generative AI depends on data quality and how it is processed and stored.

Implementing generative AI helps build the infrastructure, AI and data governance, risk management and organizational experience needed before adopting more complex AI agent systems. As companies consider deploying AI agents, they should reassess and modify their AI and data-related governance policies and guardrails to address the evolving and unique risks posed by AI agents – primarily autonomous decision-making and operational integration.

Here are four key areas for board focus as a company’s efforts move forward from generative AI to AI agents:

Full implementation of generative AI: Successful deployment with strong governance and risk management is a prerequisite and the foundation for more complex deployments of autonomous AI agents and ultimately agentic AI.

Deployment of AI agents: Deploying autonomous AI agents is a key priority because it is the building block for deploying more complex agentic AI. Staged deployment of agents enables companies to gain operational familiarity and establish the necessary data and governance frameworks so the company is better positioned to handle more complex, adaptive and autonomous AI capabilities.

ROI and the bottom line: Wherever the company is on its AI journey – piloting, deploying or exploring – ROI should remain front and center.

People: Achieving the transformational benefits of generative AI and AI agents at scale hinges on placing people at the core: prioritizing change management, workforce empowerment, skills development and cultural transformation. What skills are necessary? How do we mitigate the risk of de-skilling? It isn’t enough for generative AI and agentic AI to be deployed: they must be embraced. Fundamentally changing what people do every day and how they work will require leadership.

David Brown is executive director of KPMG’S Board Leadership Center