Artificial intelligence: What the C-suite needs to know 

Investments in artificial intelligence (AI) have grown in recent years and discussions are now shifting from how to create business value with AI to how to do so in a responsible and ethical way. Given the potential of AI for value creation, disruption and destruction, it is imperative for executives to better understand and manage "the Art of the Possible and the risks of what is possible” with this technology.

To do so, C-suite executives need to ask the right questions of themselves and their teams. They need to ask not only about what strategic options AI can create but also what can make AI risky and what types of risks they need to consider when deploying AI. They need to identify the most effective organizational design to embed AI across an organization and think about AI governance. This must include what practices organizations can use to ensure AI is deployed ethically, in compliance with new regulations and with a focus on ensuring their customers’ and other stakeholders’ trust

For complex topics such as the implications of AI for business and society, asking the right questions can be as important – or more important – than identifying the potentially many “right” answers. Moreover, getting answers to the questions AI raises requires a multi-faceted approach and holistic understanding of AI that spans technical, organizational, regulatory, societal and even philosophical factors.

This is the approach we have taken with the World Economic Forum AI C-Suite Toolkit, which has been developed based on insights from industry and academic experts. It aims to help executives guide their organizations to take advantage of this technology responsibly.

Consider the strategic implications of AI, for example. Unlike other technologies, AI has the ability to automatically (or semi-automatically) make decisions, from what products to recommend to which customers; to how to prioritize customers or even patients in medical triage cases. This creates a fundamental shift in the cost base and scalability of businesses as AI can replace people-dependent, variable costs with the fixed costs of AI-enabled software.

AI can take automation further than what was previously possible. Many tasks that are repeated frequently rely on making predictions, such as what customers are likely to default on a loan. Tasks that AI can learn to perform using data can be automated and improved

Business metrics traditionally affected by technology can also be improved with AI, from KPIs related to customer satisfaction or engagement to process efficiency metrics or employee productivity.

AI also has the potential to disrupt industries, as we can already see in industries such as automotive or insurance.

While the opportunities increase as the AI capabilities of organizations mature, the journey of an organization to AI maturity is not an easy one. Executives need to be able to assess their organization’s AI maturity and then understand the roadblocks to progress. They must consider the organizational and cultural changes that need to be made and those that will occur following the introduction of AI

Defining the appropriate organizational design with which to embed AI and data-driven decision-making across an organization is complex. Multiple challenges must be overcome to align AI with all parts of the organization, from engineering to customer-facing units, and to upskill the workforce effectively. Success – even at the level of specific AI project implementations – is not a given: executives need to understand new project execution risk factors (beyond usual ones such as change management challenges) that can lead to costly project failures. These may include very challenging data issues or difficulties from continuous risk management of the AI models deployed.

This only covers half of the AI map for executives. The other half involves the new AI risks that organizations and society face. With power comes responsibility; this is true for AI. The ability of AI-enabled systems to make automatic decisions leads to a number of new risks, from privacy infringement to potential discrimination, for example, when similar customers are treated differently by AI; to new safety and security risks.

Executives need to understand not just how their organizations can adopt AI, but how to do so responsibly. “Successful” adoption of AI may actually prove costly in the long term if these new risks materialize. Success with AI is both a matter of value creation and of risk management for the business. Executives need to answer questions such as: what are the new risks AI creates for my business? How can we manage these risks? What is the role of AI governance and how can we best think about it?

AI has been in development for more than half a century. This is how long it takes for major technological innovations to reach critical mass in terms of adoption and real impact. It is only in the past few years that we have started seeing the potential of AI to transform businesses and society.

The World Economic Forum has developed the AI C-Suite toolkit to provide a map of questions the C-suite should consider as they guide their organizations in their responsible AI journey. The questions are laid out across the AI lifecycle starting with what AI can do for the business, strategy, people and organizational change, with a particular focus on implementing responsible AI. The toolkit also provides specific questions that executives need to ask about AI and sustainability as well as about industrial AI.

In the early stages of this journey, we are only just starting to understand some of the positive and negative impacts of AI. Responsible AI innovation and adoption is not yet a given. C-suite executives must play a critical role in achieving this and need to understand multiple aspects and ask very diverse questions.