The path from a pilot project to a bank-wide artificial intelligence system is littered with failed experiments. The World Economic Forum, in a report released Tuesday in Dalian, suggests the financial industry is now trying to clear that path. The report, developed with Accenture, is called The AI Playbook for Financial Services.
It is the result of 18 months of roundtables and interviews with senior leaders from over 100 organisations. The central finding is blunt: firms are done with small tests.
They are moving to broader use. But the report also warns that the next phase is harder. Trust, governance, and human oversight are now the bottlenecks.
This is not a technology problem anymore. The technology works.
The problem is organisational. Banks and insurers have spent years building isolated AI experiments in marketing or fraud detection. Now they need to weave those systems into the core of how work gets done.
That means redesigning jobs, not just adding software. The report draws on case studies from multiple financial institutions. It does not name them, but the pattern is clear.
Firms that rushed to deploy AI without fixing their data foundations hit walls. Those that invested in governance first moved faster in the long run.
Speed and security are not opposites, the report argues. They are two sides of the same strategy. Drew Propson, Head of Technology and Innovation in Financial Services at the World Economic Forum, put it this way: the most successful organisations focus on two priorities at once.
They chase immediate gains — productivity, customer experience, operational efficiency. At the same time, they build the data, governance, and technology infrastructure needed to scale safely.
It is a balancing act, and most firms are still learning it. The stakes are high. As AI tools become cheap and widely available, a bank’s ability to compete will depend less on the algorithm and more on the quality of its customer relationships.
That is a human factor. It requires transparency. Customers need to know when AI is making a decision about their loan, their insurance premium, their credit card limit.
They need to trust that a human is still in the loop. The report does not pretend this is easy.
The roundtables revealed deep unease among executives. They worry about regulatory backlash. They worry about reputational damage if an AI system makes a bad call.
They worry about losing control of decisions that affect people’s lives. That unease is healthy.
The industry has seen what happens when tech companies deploy AI without guardrails. Financial services is different. A chatbot that gives bad shopping advice is a nuisance.
A lending algorithm that discriminates is a scandal. The margin for error is razor-thin. The report’s timing matters.
The Annual Meeting of the New Champions is a gathering of the global elite, a place where the future gets discussed in conference rooms and hotel lobbies. That the WEF chose to release this playbook now signals that AI in finance is no longer a speculative topic.
It is a live operational issue. Firms that get this right will not just save money. They will build a competitive moat.
Firms that get it wrong will face regulators, lawsuits, and angry customers. The report offers a roadmap, but it does not guarantee success.
Execution is everything.
