BoJ's Uchida sets out how AI touches monetary policy — and why the size is still unknown

Victoria Shinder17 October 2026, 09:072 min read

Bank of Japan Deputy Governor Shinichi Uchida used his opening remarks at the ECONDAT 2026 Fall Meeting, published by the BoJ (dated 5 October 2026), to lay out how artificial intelligence bears on the conduct of monetary policy. His framing is cautious: the directions of AI's effects fit within the traditional policy framework, but "to what extent and degree? In what time horizons? We don't have a clear answer yet."

AI as a central-banking tool. Uchida said AI and big data are removing two constraints on central-bank analysis — limited computational power and limited data. Machine learning and generative AI let researchers handle far larger datasets, and large language models turn unstructured data into structured data. He cited uses at the BoJ: high-frequency human-mobility data to gauge the pandemic recession in real time, and vessel-tracking data to monitor the Middle East conflict's effect on crude-oil imports and supply chains. He cautioned that the effectiveness of non-traditional data "might depend on the state of the economy and the relevance of data".

AI in the policy parameters. Calling AI a general-purpose technology that, unlike the steam engine or the internet, substitutes for cognitive rather than physical labour, Uchida said it bears on "some core parameters of monetary policymaking, including the output gap, financial conditions, and star variables":

• it is a big positive demand shock, pushing up the economy and prices;
• it could raise supply through productivity and capital accumulation, which might affect **r-star (r\*)**;
• it has boosted stock prices, easing financial conditions, while large bond issuance by AI-related companies pushes up long-term rates, tightening them;
• it may change labour markets structurally.

Each, he said, works in different directions over different horizons. Tentatively, "the demand side has come first and… has been making financial conditions more accommodative on balance, while there is a risk of correction, if profits do not follow"; the effects on r\* and u-star are "hard to gauge".

The policymaker's posture. Uchida said he asks himself whether he is making balanced decisions, "neither underestimating nor overestimating the impacts of AI", and that central banks assess the whole economy, including sectors less affected by AI. He noted conventional statistics may not keep up with AI's adoption, hoped alternative data could help, and placed the task in a longer-term frame alongside geopolitical tensions, climate change and frameworks to control AI.

These are a deputy governor's framing remarks at a research conference, not a policy decision or signal about the BoJ's rate path; the value here is the explicit list of channels and Uchida's repeated point that their magnitude and timing remain open questions.

Source: Bank of Japan, "Opening Remarks by Deputy Governor UCHIDA at the ECONDAT 2026 Fall Meeting (AI, Big Data, and Monetary Policy)" — http://www.boj.or.jp/en/about/press/koen_2026/ko261005a.htm