Dan Priest, Chief AI Officer at PricewaterhouseCoopers LLP, Jenn Kosar, AI Assurance Leader, and Barbara Berlin, Managing Director, have outlined critical actions for boards of directors to effectively govern and oversee the accelerating integration of Artificial Intelligence (AI) into business operations. Their insights, derived from a recent PwC memorandum, emphasize that AI’s rapid evolution from a breakthrough technology to a core driver of business transformation necessitates strategic foresight and disciplined execution from the highest levels of corporate governance.
Since the public debut of ChatGPT in late 2022, the landscape of AI has undergone a dramatic metamorphosis. What was once a niche area of technological exploration has rapidly become a central force reshaping how businesses operate across all facets, from product development and customer service to strategic decision-making and workforce management. The memo underscores that the ultimate success of AI integration will hinge not merely on the sophistication of the tools themselves, but on how efficiently companies redesign their entire business infrastructure around these capabilities. This profound shift demands clear strategic choices, judicious investment, adaptation of operating models, and a dedicated focus on workforce development. Crucially, building and maintaining trust through robust risk management, strong governance, and responsible AI deployment are paramount for achieving competitive advantage.
The role of the board in this dynamic environment is to ensure management is making prudent decisions, fostering the right conditions for AI success, and demonstrating the agility required to navigate the evolving technological, competitive, and regulatory landscapes. Directors, acting on behalf of shareholders, must provide short-term governance while also challenging management on the long-term strategic choices that will define the future of the enterprise.
Governing AI as a Transformation, Not Just a Tech Initiative
Effective AI governance, according to the PwC experts, begins with establishing clear accountability. As AI transitions from experimental phases to enterprise-wide transformation, boards must oversee the strategic decisions that propel this change and hold management accountable for tangible results. This requires a deep understanding of how AI impacts the company’s strategy, operating model, workforce, corporate culture, and overall risk profile.
A key recommendation is to clarify board and committee oversight responsibilities. Given AI’s direct connection to business strategy, enterprise transformation, and long-term shareholder value, the full board typically bears the ultimate oversight responsibility. However, AI’s pervasive influence necessitates a cross-committee approach, as it touches upon strategy, operations, talent, risk, financial reporting, and compliance. While board-level technology committees are increasing in number—with 17% of S&P 500 boards now having one, up from 15% in 2021—AI oversight should not be confined to a single entity. Instead, each committee must engage with the aspects of the AI transformation and associated risks that fall within its purview. For instance, the audit committee might focus on the accuracy and integrity of AI-driven financial reporting, while the risk committee would scrutinize model bias and operational risks.
Establishing clear management accountability is equally vital. While some organizations have appointed a Chief AI Officer (CAIO) to coordinate AI initiatives, others may delegate this leadership to Chief Strategy Officers or Chief Operating Officers. The critical factor is not the title, but the presence of a designated individual or team accountable for a strategy-driven, outcomes-focused AI plan. This leader must orchestrate the AI agenda, fostering collaboration among business unit heads, finance, legal, risk, human resources, and other C-suite executives. Furthermore, many companies are forming management-level AI councils to align priorities, build robust risk and governance frameworks, and drive AI adoption.
A significant challenge highlighted is the need for enhanced AI fluency within the boardroom. The rapid pace of AI development and its profound impact demand that directors possess, or have access to, the necessary skills and expertise. The PwC survey reveals that 71% of directors identify AI as the board capability most in need of strengthening. Boards must assess whether their current composition requires additional education, external expertise, or the recruitment of directors with specialized AI knowledge, tailored to the company’s specific business model, industry, and risk appetite. While directors are not expected to be AI engineers, sufficient fluency is essential for confident oversight. This fluency can be cultivated through direct engagement with AI tools, specialized board education sessions, external training programs, and even leveraging the company’s own employee AI training initiatives.
Intriguingly, the use of AI within the boardroom itself is emerging as a component of effective oversight. While current adoption remains modest—with only 40% of directors reporting AI use in their oversight roles—boards can leverage AI for strategic thought partnership, synthesizing complex board materials, conducting research and scenario modeling, and preparing more efficiently for meetings. The key principle is that AI should augment, not replace, human judgment, and its application must be accompanied by a thorough understanding and management of confidentiality and legal risks. A proactive governance framework is recommended for responsibly integrating AI into board processes.
Aligning on Strategic Priorities: Leading, Lagging, or Exiting with AI
As AI investment escalates, boards must oversee how this technology serves as an accelerator for business strategy and long-term shareholder value, rather than viewing AI as a strategy in itself. The PwC memorandum stresses that AI is a tool, not an end goal, and boards should challenge management to articulate where the company intends to lead with AI, where it aims to keep pace, and where it might strategically choose to exit or deprioritize initiatives based on tangible business outcomes. This requires a clear dialogue about the company’s current AI maturity, realistic sector-specific possibilities, and the most promising avenues for value creation.
The reality is that not all AI ventures yield immediate returns. While some companies are realizing significant revenue gains and cost savings, many others remain mired in experimentation. The differentiator for companies achieving a strong return on investment lies not in the sheer volume of AI activity, but in focused investment aligned with strategic priorities, rather than scattered, low-impact experiments.
Boards should expect management to provide insights into industry evolution, potential disruptors, and the company’s competitive AI posture over time. Management must make explicit choices about the company’s AI ambition: where to lead with bold initiatives, where to maintain pace, and where to strategically disengage. These choices should reflect the maturity of the underlying AI technologies. While AI has an expansive appetite for use cases, not every application translates into meaningful business outcomes. Some AI capabilities are mature and scalable, while others are nascent and require further investment and experimentation before delivering value. Boards must encourage management to "pick its spots" and align investment with AI’s realistic potential to create value.
To successfully lead in AI, companies must establish five core enablers:
- Data Readiness: Data quality, usage rights, consent, security, and governance are critical gating factors. Poor data practices can hinder AI initiatives and amplify operational, legal, privacy, and reputational risks.
- Technology and Infrastructure: Major technology decisions regarding AI models, cloud computing, proprietary versus third-party tools, and data architecture significantly influence speed, cost, risk management, vendor dependency, and long-term resilience. Companies are increasingly seeking to maintain flexibility by avoiding over-reliance on single models or vendors.
- Talent and Capabilities: The ability to attract, develop, and retain AI talent, along with fostering AI fluency across the workforce, is essential for creating lasting business value.
- Operating Model and Governance: Redesigning work, team structures, decision rights, and accountability frameworks to integrate AI agents and human judgment effectively is crucial for scalable AI adoption.
- Risk Management and Controls: Establishing a robust AI risk management foundation, including policies, an inventory of AI use cases, a common risk taxonomy, and embedded controls, is vital for scaling AI safely and building stakeholder trust.
The PwC benchmarking data from 2022-2025 (excluding AI foundation builders like Alphabet, Microsoft, and NVIDIA) reveals a widening AI performance gap. Companies investing over 0.5% of revenue in AI outperformed their sector median total shareholder return by 21%, while those investing less underperformed by 2%. Similar trends are observed in operating performance, with higher AI investors exceeding sector median revenue growth by 3% and EBITDA growth by 7%. This underscores that companies treating AI as an enterprise-wide growth and transformation lever are outpacing those still relying on isolated pilots and low-risk automation.
Steering Talent and Culture for AI Adoption
AI transformation is fundamentally a people-centric endeavor. While AI capabilities are becoming more accessible, sustained business value creation increasingly depends on having the right leaders, skills, and culture. Boards must ensure management possesses a comprehensive talent strategy, robust organizational capabilities, and effective change management processes to embed AI across the enterprise.
This necessitates increased engagement with the Chief Human Resources Officer (CHRO), alongside business and technology leaders. Discussions should extend beyond basic workforce planning to encompass leadership development, critical AI skills acquisition, workforce readiness, and the cultivation of a culture that promotes learning, responsible experimentation, and continuous adaptation.
The companies deriving the most value from AI are those with the right personnel guiding, designing, governing, testing, and adopting AI-enabled workflows. Human capability—the ability to reimagine work, make informed decisions, manage risks, and drive adoption at scale—will be a key differentiator. Boards should inquire about management’s plans for developing leaders, attracting and retaining top AI talent, and equipping the broader workforce with the skills to thrive in an AI-augmented environment.
Fostering a culture where AI adoption takes root requires employee trust. Employees are more inclined to embrace AI when they trust leadership, understand the rationale behind changes, and feel supported in acquiring new skills. Management must implement change management programs that include leadership alignment, targeted upskilling and reskilling, transparent communication, and clear accountability for AI adoption. Addressing common adoption barriers, such as change fatigue, unfamiliarity with AI, concerns about job displacement, and uncertainty regarding evolving roles, is also critical.
Measuring the success of AI adoption should focus on tangible business outcomes, not just activity metrics. Metrics like usage rates, user numbers, or tokens consumed may indicate experimentation but not necessarily value creation. Companies should pair usage data with outcome-oriented indicators, such as employee engagement, workforce sentiment, and progress against strategic goals. This holistic approach enables boards to assess AI’s integration into daily operations and its impact on workforce effectiveness and overall enterprise performance.
Guiding the Workforce of People and AI Agents
As companies integrate AI into core workflows and transition to a hybrid workforce of humans and AI agents, existing talent and operating models face significant disruption. The most successful organizations approach agentic AI adoption as a holistic workforce and operating model transformation, not merely a technological rollout. This involves redesigning work processes rather than simply automating existing tasks.
AI agents can proficiently handle routine inquiries, perform complex reasoning, execute multi-step processes, generate code, and amplify the capabilities of knowledge workers. However, the true value emerges from a human-led, tech-powered model where people guide, oversee, and collaborate with these agents, combining AI’s analytical power with essential human judgment. Boards should expect management to proactively rethink end-to-end workflows, team structures, decision rights, and accountability, rather than simply appending AI to existing complexities.
Crucially, human judgment must remain at the core of decision-making. As agents take on more execution-oriented tasks, human roles will increasingly focus on judgment, exception handling, and risk oversight, with humans retaining ultimate accountability for outcomes. Management must clearly define areas where human judgment is non-negotiable, especially for high-impact decisions. This includes establishing protocols for when AI outputs require human approval, review, or override, and ensuring AI outputs are explainable and auditable within the relevant context.
Directors should expect management to articulate how operating models and talent strategies will adapt to accommodate a workforce comprising both people and AI agents. Management should demonstrate how work will be redesigned, encompassing decision rights, accountability, governance, team structures, and the development of critical skills as AI agents become integrated into core business processes. Companies must also reassess their sourcing models, determining whether to bring AI and related talent in-house or maintain outsourcing arrangements based on evolving capabilities.
An illustrative example in customer service highlights this shift. In traditional models, humans handle most interactions across tiered support. With AI agents, a significant portion of routine tasks—answering common questions, summarizing customer histories, drafting responses, and processing simple requests—can be automated. Human employees then transition to higher-value tasks like resolving complex exceptions, managing sensitive situations, enhancing service quality, and intervening when agents err or exhibit uncertainty. This necessitates a workforce with new skills, including those who design, train, test, and improve AI agents, working with prompts, data, and controls. Experienced customer service professionals become even more valuable for their understanding of customer needs, service standards, and good judgment, evolving from task execution to supervising AI and optimizing human-AI collaboration. For boards, the fundamental takeaway is that this is not mere automation; it is a comprehensive redesign of roles, skills, and accountability across the entire service model.
Overseeing AI Risks and Controls for Scalable Adoption
Transforming with AI while maintaining stakeholder trust is not an ancillary compliance function; it is the foundational enabler for safe and rapid AI scaling. The challenge for directors is to ensure AI risks are comprehensively understood, appropriately prioritized, and effectively managed without stifling innovation. Robust governance over AI risks should empower companies to move forward with greater confidence and bolster trust among all stakeholders.
Boards should begin by confirming that management has established a clear AI risk management framework. This includes implementing clear policies, maintaining an up-to-date inventory of AI applications across the enterprise, developing a common risk taxonomy, and defining baseline controls. This inventory should encompass AI developed internally, AI embedded in third-party software, and AI tools utilized by employees. AI risks must be seamlessly integrated into the company’s broader risk management approach, employing a shared language for assessing exposures across model, data, infrastructure, and user risks, alongside legal, compliance, and process impact risks. Human oversight remains critical, and management must clearly delineate ownership of AI risk decisions, models, escalation procedures, and remediation processes. Given the rapid pace of AI advancement, emerging risks may exceed the capacity of current risk management structures, necessitating ongoing updates to these frameworks. Controls should be embedded into AI initiatives from their inception, spanning design, testing, deployment, and post-deployment phases.
A key risk-tiering approach is recommended, where governance processes are differentiated based on AI risk levels. Streamlined procedures can be applied to lower-risk uses, while higher-risk systems require more intensive review, robust controls, clear escalation pathways, and heightened oversight. Risk tiers should be determined by objective criteria, considering factors such as sensitive data usage, external exposure, autonomous decision-making, potential business impact, and regulatory sensitivity. Boards should receive targeted reporting on the highest-risk models and use cases that present the greatest potential exposure, particularly those involving significant judgments, decisions, or the potential for high consequences in case of failure, drift, or misuse. Independent validation or assurance by internal audit or third parties may be appropriate for these critical models.
Management must implement continuous monitoring of AI-enabled workflows to detect incidents, control breakdowns, and other emerging risks. Continuous monitoring is essential as AI systems can exhibit drift, unpredictable behavior, generate erroneous or harmful outputs, or degrade over time. Boards must be promptly informed of significant incidents, with pre-defined escalation triggers and updated resilience and response playbooks tailored for potential AI-related events.
The regulatory environment for AI is still developing, presenting a fragmented landscape. In the U.S., the federal approach generally favors innovation, but a comprehensive federal AI law is absent. Companies must navigate a mix of agency guidance, existing regulations, and state-level requirements. The White House’s National Policy Framework for AI, released in March 2026, aims to preempt state-level AI regulations, preventing a complex patchwork of laws. Globally, numerous regulatory initiatives are underway, with the landscape continuously evolving. Boards should inquire about management’s proactive monitoring of these regulatory developments and their strategies for building adaptable governance frameworks.
Monitoring Outcomes and Risks for Sustained AI Value
The board’s fundamental role is to monitor whether management’s AI initiatives, investments, and operating model adjustments are yielding tangible business value. This includes assessing capital allocation discipline over time and identifying emerging risks. Boards should receive consistent reporting on the AI transformation journey, encompassing progress against goals and milestones, challenges encountered, risk management effectiveness, and the need for strategic pivots.
As AI transformation scales, reporting must transcend mere activity updates. It needs to demonstrate value creation, highlight emerging risks, and confirm the establishment of conditions for sustained success. While many companies are in the early stages of AI adoption and may not yet possess a holistic set of value-defining metrics, these metrics become increasingly critical as AI integration expands.
Current AI reporting often falls short, with few directors receiving quality information linking AI investments to business performance. Boards must understand who owns AI strategy, execution, and governance, and the frequency of engagement between key leaders and the board. Joint presentations by the CAIO or CIO with business or functional leaders can help ensure alignment between technological efforts and desired business outcomes.
Measuring AI ROI can be complex, with each company defining its own success criteria. Nevertheless, directors should expect clear, relevant metrics tied directly to the transformation’s strategic goals. Depending on the industry, these metrics might include efficiency gains, cost savings, incremental revenue, enhanced quality, cycle-time improvements, and elevated customer experiences. Metrics focused solely on AI usage, model counts, or tool adoption are less impactful for ROI assessment.
Building stakeholder trust through transparency is paramount. Clear communication about AI usage, governance, and risk mitigation fosters confidence both internally and externally. Directors should understand how management demonstrates responsible AI use and cultivates trust with investors, regulators, employees, and other stakeholders. Attention to public messaging is also important, ensuring it is evidence-based, avoids hyperbole, and guards against "AI washing." Boards must be prepared to address investor inquiries regarding AI oversight, the linkage of AI initiatives to business strategy and outcomes, the locus of oversight responsibilities, and the board’s possession of requisite skills and expertise.
Conclusion
The potential for AI to generate significant value is substantial, as are its inherent risks. However, realizing AI’s transformative effects necessitates focused investment and disciplined execution. Boards play a pivotal role in guiding management to approach AI as an enterprise-wide transformation and providing critical oversight of the decisions that will shape the future of their organizations.
