The latest report from the Capgemini Research Institute, Inside the C-Suite: How AI is quietly reshaping decisions, shows that more than half of C-level executives now use AI to support their decision-making processes.
Key figures from the report
According to the study, 92% of active AI users in executive leadership teams rely on it every day. Meanwhile, 39% of leaders follow a partnership model, working with AI to assess the basis of recommendations and arrive at a final decision together. One in six senior executives already uses algorithms at every stage of strategy development, and that share is expected to rise to 35% within the next three years. At the same time, only 11% of respondents openly admit that AI tools are part of their decision-making process.
How executive teams are using new technologies
The use of AI across leadership structures is continuing to expand. Today, on average, one in six top executives uses algorithms throughout the full strategy development process. Within three years, that figure is projected to climb to 35%. For many leaders, AI has already become part of their daily routine, with 92% of those who have implemented it using it every single day.
The report also highlights important differences in technology adoption at the executive level. Chief technology officers are the most frequent users of AI, with 55% saying they use it in decision-making. However, chief executive officers are also emerging as an especially active group. Already, 41% of CEOs use AI in decision-making processes, the highest figure among all non-technical leadership roles. Notably, CEOs use algorithms almost twice as often as other board members outside IT functions.
“Today, the term AI in the boardroom covers a broad spectrum of decision-support technologies. We can see that leaders are becoming increasingly skilled at combining traditional models with the capabilities of generative AI. CEOs are currently the group most willing to experiment with these tools, and their adoption rate is rising the fastest. The next and most promising step in this evolution is the rise of AI agents. They allow actions to be coordinated autonomously across multiple systems at the same time, making it possible, for example, to monitor performance on an ongoing basis and recommend strategic shifts in resources in real time,” says Maciej Sowa, Principal Delivery Head, BSv at Capgemini Poland.
What CEOs are asking algorithms
An analysis of prompts sent to the most popular AI tools shows that as many as 58% of work-related queries concern two key areas. The first is the acquisition, documentation and interpretation of information. The second is support in decision-making, advice, and creative problem-solving. In addition, 24% of all work-related prompts relate to writing and editing tasks, while 15% focus on seeking practical guidance.
“Although AI currently supports day-to-day productivity above all, leaders expect that in the near future the technology will become a partner capable of challenging their existing strategic thinking,” notes Maciej Sowa, Principal Delivery Head, BSv at Capgemini Poland.
Executives are aware of AI’s pitfalls
Although nearly half of leaders declare a high level of trust in AI’s role in decision-making, the way they use it suggests considerable awareness and caution. A partnership-based and iterative approach dominates at the top of organizations. Some 39% of executives prefer to develop decisions in cooperation with AI, while 17% use it to re-check analyses. Only 1% of decision-makers admitted to following recommendations uncritically. This caution stems not only from concerns about data errors or legal risks, but also from phenomena such as “persuasion bombing.”
A revolution behind closed doors
Perhaps the most striking conclusion of the report is that this shift in decision-making is taking place behind closed doors. Although AI is already playing a real role in shaping strategy, leaders still feel uncomfortable disclosing that fact, with only 11% speaking openly about it. These concerns stem in part from the inability to fully reconstruct how algorithms operate and therefore document the decision-making process, as well as from the potential reactions of the wider business environment.
“The conversational style of artificial intelligence is extremely persuasive. There is a risk that while a user may begin a conversation with AI to verify facts, the virtual assistant may shift into a mode where it tries to persuade them toward a particular point of view,” says Maciej Sowa, Principal Delivery Head, BSv at Capgemini Poland.
The report also points to an important gap. Leaders are using AI extensively as individual users, for example to analyze data or draft emails, but their organizations still struggle to implement AI systematically across business processes. One of the key priorities, therefore, should be the creation of a corporate governance framework that brings structure and clarity to how AI is used within the organization.
