AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Czech National Bank Governor Aleš Michl said the bank uses AI to forecast inflation and GDP, and that its current-quarter GDP forecast performs about as well as forecasts by economists. The speech describes the model and its data inputs, but provides no comparative figures or detailed evaluation method.

The Czech National Bank uses artificial intelligence to forecast GDP, and its current-quarter estimate performs about as well as forecasts by economists, Governor Aleš Michl said in a lecture at Kyoto University on November 13. The comparison points to a possible role for AI in economic forecasting, but the speech did not include accuracy figures or explain how the forecasts were evaluated.

Before you orderOffer from Amazon

Get smart everyday buys delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

Michl said the bank began developing short-term GDP forecasting methods during the COVID-19 pandemic, when conventional economic data were blurred and delayed. The central bank started tracking faster-moving indicators, including electricity consumption and truck traffic on highways, and combined them into a weekly index.

In 2023, the bank added machine learning to the work, Michl said. He described the system as one tool among the bank’s forecasting methods, saying it helps predict both inflation and GDP. His example focused on forecasting GDP for the current quarter, when official national accounts data may not yet provide a complete picture.

The governor’s statement that the AI forecast performs as well as economists is a claim made in the lecture. The speech did not specify which economists or forecasts were used as comparators, the period covered, the error measure, or whether the comparison was independently assessed. It also did not say that AI has replaced the bank’s economists.

At a glance
reportWhen: Lecture delivered November 13, 2026; re…
The developmentIn a November 13 lecture at Kyoto University, Czech National Bank Governor Aleš Michl said the bank’s AI-assisted current-quarter GDP forecast performs about as well as economists’ forecasts.

Faster Signals for GDP Forecasts

GDP figures are released after the economic activity they describe, while central banks must make decisions using information available at the time. A model drawing on weekly indicators such as power use and road freight could help policymakers gauge shifts before full quarterly data arrive. That may be useful when assessing the outlook for output and inflation.

The comparison with economists matters because it bears on whether machine-learning tools can support a core central-bank task. If a model can produce estimates as accurate as human forecasts, it could become another input for analysts and policymakers. But Michl’s description alone does not establish that result across different periods or economic conditions. Forecast performance depends on the data, method and evaluation window, none of which were detailed in the speech.

Products Worth Considering

Dell XPS Tower 8960 Desktop, Intel Core i7-13700, NVIDIA GeForce RTX 4060, 16GB DDR5, 512GB SSD

Dell XPS Tower 8960 Desktop, Intel Core i7-13700, NVIDIA GeForce RTX 4060, 16GB DDR5, 512GB SSD

As an affiliate, we earn on qualifying purchases.

From Pandemic Data to Machine Learning

The bank’s account traces its short-term forecasting work to the disruption caused by the pandemic. In periods when standard statistics arrive late or are difficult to interpret, more frequent indicators can give analysts an earlier, though not necessarily complete, view of economic activity. Electricity demand and truck movements are examples of data that can change more quickly than quarterly GDP estimates.

Michl presented the AI work as part of a broader technology program at the Czech National Bank. He said the bank runs AI in a hybrid setup: its own servers and open-source models for sensitive information, alongside cloud services for other uses. The GDP forecasting example is separate from the bank’s experimental work on digital assets and distributed-ledger technology, which he also discussed in the lecture.

How the Forecast Comparison Was Made

The lecture did not provide forecast error statistics, a sample period, or a description of the benchmark used to support the claim that the AI forecast performs about as well as economists. It is unclear whether the comparison concerns predictions made in real time, later revised data, a particular group of economists, or a specific forecasting exercise.

Michl also did not explain the model’s full range of inputs, how often it is updated, or how the central bank weighs its output against other evidence. The speech gives no indication that the model determines monetary policy or that its forecasts have been independently validated. The claim should therefore be read as the governor’s account of the system, not as a published head-to-head evaluation.

More Evidence Needed on Accuracy

The next useful evidence would be a technical account from the Czech National Bank describing the model, its data and how its forecasts compare with economists’ estimates over a stated period. Results across multiple quarters, including periods of rapid change, would help readers judge how reliably the system performs.

Michl’s lecture did not announce a new deployment date or a change to the bank’s forecasting process. For now, the confirmed development is that the central bank says it uses machine learning in short-term GDP forecasting and considers its current-quarter performance comparable to economists’ forecasts. Whether the bank publishes supporting results or expands the model’s role remains unclear.

Key Questions

What did the Czech National Bank say about AI and GDP forecasts?

Governor Aleš Michl said the bank uses AI to forecast GDP and that its current-quarter forecast performs about as well as economists’ forecasts. The lecture did not provide figures to measure that comparison.

What data does the bank use for its short-term GDP forecasting?

Michl cited electricity consumption and truck traffic on highways as fast-moving indicators used to build a weekly index. He did not list every input to the model.

When did the Czech National Bank add machine learning?

Michl said the bank added machine learning to its short-term GDP forecasting work in 2023. He traced the development of the underlying methods to the COVID-19 pandemic.

Has the bank shown that its AI forecast is more accurate than economists’ forecasts?

No such evidence was included in the lecture. Michl said the forecast performs about as well as economists’ forecasts, but gave no comparison method, accuracy figures or evaluation period.

Source: bis.org

This content is for general information only and is not financial, tax or legal advice. Consult a qualified professional for decisions about your money.
You May Also Like

Sainsburys Argos Sale

Sainsbury’s has announced the sale of Argos to a private equity firm, marking a significant shift in its retail strategy. The deal is valued at several billion pounds.

Christine Lagarde: Interview With Les ÉChos

ECB President Christine Lagarde outlines monetary policy outlook and economic challenges in a detailed interview with Les Échos.

Visa, Mastercard, Major Banks Facing New Litigation Over ‘Anticompetitive’ Fees

A San Diego pizzeria’s proposed class action alleges Visa, Mastercard and five banks maintained restraints that kept merchant card fees high.

55 Or Older? A Veteran? First Responder? These T-Mobile Discounts Could Lower Your Cell Phone Bill

T-Mobile offers discounted plans for customers 55+, military members, veterans and first responders. Prices, eligibility and benefits vary by plan.