The AI Returns Problem Is Not a Technology Problem. It Is a Leadership Problem

PwC's 2026 research found that the organisations generating real AI value are not deploying better tools — they are making better decisions. That distinction begins and ends in the boardroom.

Video summary: https://youtu.be/pObx3p_FSeY

Sign up for my Substack daily AI newsletter here.

See my AI Training course portfolio for corporate Business Leaders here.

Article link: https://open.substack.com/pub/johanosteyn/p/the-ai-returns-problem-is-not-a-technology?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

There is a number that every South African business leader needs to sit with before their next board meeting. It comes from PwC’s 2026 AI Performance Study, released on 13 April 2026, based on interviews with 1,217 senior executives across 25 sectors worldwide. The number is this: 74 per cent of all AI economic value is being captured by just 20 per cent of organisations. The remaining 80 per cent — the majority of businesses investing in AI, building AI strategies, appointing AI leads, and reporting AI initiatives to their boards — are sharing the other 26 per cent between them. They are not failing because of inferior technology. They are not failing because AI does not work. They are failing because of the decisions their leaders are making about what AI is for, how it should be governed, and what success looks like. That is a leadership problem. And in most organisations, it is a problem that lives directly in the boardroom.

CONTEXT AND BACKGROUND

PwC’s study is one of the most comprehensive examinations of AI performance yet conducted, analysing the impact of 60 AI management and investment practices across organisations globally. Its findings are unambiguous about where the divide originates. The top-performing 20% of companies are not simply deploying more AI tools. They are generating 7.2 times more AI-driven revenue and efficiency gains than the average competitor. And the reason, according to PwC, is not technological sophistication. It is strategic intent. The organisations capturing the most AI value are using it as a catalyst for growth and business reinvention — pursuing new revenue opportunities, crossing traditional industry boundaries, and redesigning business models rather than simply making existing operations marginally more efficient. The majority are doing the opposite: pointing AI at cost reduction and efficiency within existing business lines, and wondering why the returns do not materialise.

PwC Global Chairman Mohamed Kande put the situation plainly: a small group of companies are already turning AI into measurable financial returns, while many others are still struggling to move beyond pilots. The research identifies the specific mechanism behind the divide. Industry convergence — using AI to expand beyond traditional sector boundaries and pursue growth opportunities created as industries intersect — is the single strongest factor influencing AI-driven financial performance, ahead of efficiency gains alone. Companies in the top group are two to three times more likely than their peers to use AI to identify and pursue these cross-sector growth opportunities. They are also 2.6 times more likely to report that AI improves their ability to reinvent their business model. These are not technology statistics. They are strategy and leadership statistics.

INSIGHT AND ANALYSIS

The governance dimension of this divide is equally striking. PwC’s research found that the leading organisations have built what it calls trust at scale — structured frameworks including responsible AI governance boards, which they are 1.5 times more likely to have, and formal responsible AI frameworks, which they are 1.7 times more likely to have implemented. The payoff is measurable: employees at leading companies are twice as likely to trust AI outputs, creating a virtuous cycle in which trust enables automation, which enables further scaling. This is a crucial finding. The organisations generating the most AI value have not just pointed AI at growth. They have built the governance infrastructure that makes scaling AI safe and credible — and that infrastructure was a leadership decision, not a technology decision.

Grant Thornton’s 2026 AI Impact Survey, based on 950 C-suite and senior business leaders, reinforces exactly this point. It found that 78 per cent of business executives lack strong confidence that they could pass an independent AI governance audit within 90 days. Organisations with fully integrated AI are nearly four times more likely to report revenue growth than those still piloting — 58 per cent versus 15 per cent. The difference, Grant Thornton is explicit, is not just technology. It is accountability. The leading organisations can show how their AI makes decisions, who owns the outcomes, and what happens when something goes wrong. For the majority, that accountability is entirely absent.

The pattern that emerges from both studies is consistent and uncomfortable. Most organisations have an AI activity problem presenting as an AI strategy problem. They have pilots, reports, initiatives, and roadmaps. What they do not have is a clear answer to the questions that determine whether any of that activity converts into value: What business problem are we solving? What measurable outcome are we targeting? Who is accountable for delivering it? What will we do differently if it does not materialise?

As CIO Magazine’s analysis of what boards need from technology leaders in 2026 makes clear, the world does not need more AI pilots, more automated workflows, or more isolated proofs of concept. It needs enterprise leaders who can see the AI capability they are building clearly, govern it decisively, measure it rigorously, and articulate it with the clarity that boards and shareholders deserve.

IMPLICATIONS

For South African business leaders, the PwC findings carry particular urgency. In a market characterised by constrained capital, subdued growth, high operating costs, and intense pressure on margins, the gap between AI investment and AI return is not an abstract strategic inconvenience. It is a material financial risk. Every rand spent on AI activity that does not convert into measurable business value is a rand not spent on the people, the training, the governance, and the change management that would actually move an organisation from the 80% to the 20%. And as PwC is explicit, without a shift in approach, the performance gap between AI leaders and laggards will continue to widen as leading companies learn faster, scale proven use cases, and automate decisions safely at scale.

The practical implication for boards is equally clear. AI governance is no longer a technology committee agenda item. It is a board-level accountability question that the Directors Institute of South Africa and governance frameworks globally are now treating as a core director responsibility. Boards need to be asking — and demanding credible answers to — a small number of questions that most are currently not asking at all. What is our organisation’s AI fitness index, and how does it compare to our sector peers? What percentage of our AI investment is directed at growth versus efficiency? What measurable financial returns have our AI initiatives produced, and how do we know? Who in our executive team is accountable for those returns? If the answers to those questions are unclear, incomplete, or absent, the organisation is almost certainly in the 80 per cent — and drifting further from the 20% with every quarter that passes without an honest reckoning.

CLOSING TAKEAWAY

The AI divide that PwC has documented is not going to correct itself through better technology, more investment, or the passage of time. It is going to widen — because the organisations already capturing 74% of AI’s economic value are using those returns to learn faster, invest more intelligently, and build competitive advantages that compound over time. For South African business leaders, the question is not whether to have an AI strategy. Almost every organisation of any size already has one. The question is whether that strategy is directing AI at growth or merely at efficiency — and whether the board is holding leadership accountable for the difference. The 74/20 split is not a market statistic. It is a leadership assessment. And for most South African boardrooms, the result of that assessment has not yet been honestly confronted.

Author Bio: Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net