Speech

Opening Remarks at the 12th IMF Statistical Forum: Measuring the Implications of AI on the Economy

November 20, 2024

Good morning, and welcome. Let me start by thanking Bert and the Statistics Department for organizing the 12th IMF Statistical Forum on ‘Measuring the Implications of AI on the Economy.’ You always choose a topic that is timely and important—and also fascinating!

It has been just two years since generative AI emerged from the lab and became a tool that anyone with internet access can use. We still feel the excitement of something new and world changing. At the same time, we are all concerned about potential harms.

AI has huge potential to boost growth and efficiency—call centers, for example, have reported productivity gains of 34 percent among new and lower-skilled workers. But AI could also disrupt labor and financial markets. And it could deepen inequality within and among countries, destabilizing societies at a time when many are already very polarized. 

To make AI a force for good that boosts inclusive economic growth, we need concerted, coordinated actions by governments, the private sector, and civil society.

And what do we need to inform those actions? Data! Reliable, timely, accurate, actionable data.

For the next two days, you will be exploring both how to measure AI’s impact on the economy, and how AI can help us do our jobs better.

Let me pose a few questions to help guide your discussions:

First, how do we strengthen our statistical systems and frameworks so that countries can better measure the impact of AI on sectors like healthcare, finance, and manufacturing? How do ensure that we capture AI investment and its impact on productivity?

Second, over many years, countries have worked with international organizations such as the IMF to build ethical, strong, and transparent standards and guidelines for collecting, compiling, and disseminating official internationally comparable statistics—all while taking into account privacy concerns.

Are the existing standards and guidelines fit for purpose in an era of widespread adoption of AI?

Third, how can we leverage AI in the production of statistics? We see central banks exploring how GenAI can help them sift petabytes of banking data to better monitor risks and refine forecasts. We see statistical organizations use AI to help classify transactions, transform unstructured data into robust indicators, and improve the accessibility of economic data. 

The IMF Statistics Department has recently established the IMF Big Data Center dedicated to supporting the use of Big Data and statistical innovation and has also led the development of StatGPT, an AI assistant that lets users talk to data, significantly reducing the time needed to find, retrieve and use data.

But those are just a few examples. We want to know: How are all of you applying this technology? How can we act together to create and share better analytical tools with policymakers in all countries? 

With better data and analysis, policymakers can make better decisions, react more quickly in a crisis, and ultimately design policies that support inclusive growth.

AI has transformative potential for the world economy and for our work—if we can figure out how to measure and use it well. I’m sure you are all up for the challenge!

Thank you.