CEOs and CFOs, take note: The 3 pillars of successful GenAI adoption

By 2026, over 80% of organizations will have used generative artificial intelligence (GenAI) in interfaces or applications running in production environments – an exponential increase from less than 5% in 2023, according to analysis by Gartner. GenAI will, therefore, dominate C-suite discussions over the coming years, with investors and shareholders demanding use of the technology as soon as possible.

However, currently most AI initiatives do not succeed, with some projections suggesting a failure rate up to 80%. Moreover, even more GenAI applications are likely to fail, given the tech's novelty and the lack of experience using it since its inception. Trusting these figures, we could extrapolate that by 2026, about 20% of organizations will not have used any GenAI, about 64% will have used it and failed and just 16% of them will have used it and succeeded.

So, how should business leaders approach adopting GenAI? And how can they maximize the probability of a positive return on investment for investors and shareholders?

From our experience collaborating with companies using GenAI in pharma, healthcare, energy and heavy industry, we know that successfully using AI and GenAI requires getting it right on three elements: value, data and people.

The question about value is deceptively simple: "Which financial value does our organization aim to generate with this GenAI product or service?" We say that it is deceptively simple because it is easy to launch initiatives to use GenAI for the sake of using it without assessing the financial value this aims to create.

Following this flawed approach equates to "finding a problem for a solution," which is what many organizations have tried to do with blockchain technology. Hence, the first step for C-suite leaders is to identify their organization's real problems, asses the financial value that solving that problem would bring to the organization and then search for potential solutions to that problem.

Examples of financially valuable uses of GenAI include analyzing competitors' public documents to find insights that would otherwise remain hidden or handling customers via chat services.

GenAI solutions – as any AI solution – are based on algorithms trained on data. The data enables the algorithms to recognize patterns and make predictions. Hence, an AI model of any kind (including a GenAI model) requires representative data related to the problem the business leader is trying to solve.

Privacy concerns may emerge at this stage but several providers offer solutions that allow safe, internal use of proprietary data to train proprietary AI. This is how the European Stability Mechanism uses GenAI-driven for idea generation with proprietary data.

Business leaders should determine if they have the necessary data for deploying GenAI models. If not, they need to plan for data acquisition while understanding that progress without it may not be successful. If data suffices, the next focus is on the third question about people.

People's perceptions of AI are highly heterogeneous and range from pure excitement to utmost fear. Business leaders should consider whether the GenAI solution they are developing will complement or substitute the people in their organization.

Without entering into a discussion about revenue augmentation versus cost reduction or what the "right" thing to do is, business leaders must decide whether they want to do what they are doing with fewer resources and thus substitute some of their people with the GenAI system or to do more of what they are doing with the same resources, enhancing the people resources that they have.

Unsurprisingly, organizations going down the road of substitution will face internal opposition, something that could backfire and compromise the otherwise successful adoption of a GenAI solution.

The journey towards successful GenAI adoption presents a significant challenge, with only relatively few enterprises expected to use this technology effectively.

Even if pursuing GenAI does not result in the intended objective, it may garner useful learnings for the organization that will foster later successes. However, addressing the questions around financial value, key data and people enablers can contribute to a more straightforward path to the successful adoption of this new technology.

Discussions about AI must land in the boardroom today. Business leaders who initiate the conversation and ask the right questions will lead the future of AI; those who don't will fall behind the curve. The time to act is now.