When polled, CEOs of large corporations have cited balancing rapid AI adoption with holistic oversight across governance, ethics and expertise gaps
Based on a combination of quantitative survey data of 213 CEO* and qualitative insights from 20 in-depth CEO interviews^ (all from firms with annual revenues exceeding US$1bn) on how respondents were navigating AI adoption and challenges in various functional areas, a global management consulting firm has published some findings.
First, 89% of respondents indicated recognizing the strategic importance of AI, with one in four indicating they felt fully prepared to integrate it at scale.
Second, 59% of respondents in firms deemed “high performing” had cited maintaining direct oversight of AI strategy, compared to 92% of respondents in firms deemed to be struggling with AI adoption.
Other findings
Some 48% of respondents in “leading firms” had cited rigorously tracking AI returns on investment, while the survey report assumes that the remainder “still prioritize short-term gains over long-term value creation.” Also:
- More than 60% of respondents had cited fragmented data and outdated infrastructure as key barriers preventing AI from scaling effectively in their organization
- 57% indicated still lacking sufficient internal expertise to meet current AI needs
- 22% of respondents running organizations with AI governance councils had indicated they consistent trackbias-detection metrics as a aspect of oversight to curb risks such as bias, regulatory pitfalls, and ethical missteps
- Qualitative interviews indicated respondents’ views that that many frontline and senior staff fear displacements, especially in labor-intensive roles. Transparency and focused upskilling seemed to assuage these worries and helped leaders avoid misallocated resources on low-impact changes, and protect current advantages against AI-driven shifts in the market
- In the interviews, respondents had highlighted just how far and fast they plan to push AI-based automation. One intended to replace their entire core business with AI. Others expected tripling digital staff, while one security industry CEO had described a proactive AI model that merges technology and personnel to outpace threats. Retail CEOs cited anticipating self-service stores, and fashion leaders had expected AI to cut development time through automated fitting and grading. Others preferred AI as an enhancer rather than a wholesale replacement, while consumer packaged goods CEOs indicated a focus on optimizing core processes. Most interviewed CEOs indicated their belief that agentic AI would reshape how business decisions are made rather than just automate processes.
The report by Kearney has recommended that CEOs should “oversee AI with Integrity, evolve with the times” by forming an AI governance council, integrating risk assessments, and regularly reviewing frameworks to head off reputational threats and ensure scalability. “Proactive oversight safeguards credibility, upholds stakeholder trust, and keeps AI aligned with business imperatives — even as regulations and technologies evolve. This forward-looking structure lets organizations adapt swiftly, maintaining both ethical rigor and the agility to seize new AI opportunities.”
*The survey respondents held regional or global CEO positions in North America (32%), Europe (28%), the Asia Pacific region (16%), Latin America (12%), the Middle East (6%) and Africa (6%) — across a wide range of industries, including financial services and IT (12%), manufacturing and industrial/online retail (8%), and automotive (7%) and energy and utilities (6%). Latin America and other regions were included to provide a global perspective. The survey employed a structured questionnaire with Likert-scale, multiple-choice, and open-ended questions, generating statistically significant data across key them.
^ Collectively, the CEOs were selected from organizations with revenues of US$1bn or more: large enterprises that are well-positioned to invest in and scale AI initiatives. These companies span a range of organizational ages: 53% were more than 10 years old; 29% were 8–10 years old, 14% were 6–7 years old, 4% were 3–5 years old, and none was less than 3 years old