Can AI automate 90% of central bank monetary policy decisions using ai that simulates global economic ecosystems in real time ?

What would it mean to hand over 90 % of central-bank policy-making to AI systems that rebuild the global economy in real time? Several central banks are already testing models that read millions of data streams and spit out interest-rate choices without human input, but today’s systems still fall short of fully autonomous rule.

Background

The Bank for International Settlements (BIS) and multiple central banks have begun piloting AI models that ingest millions of daily data streams—including trade flows, labour reports, climate anomalies, and AI-generated news sentiment—and run millions of monetary-policy scenarios per second. These pilot systems recommend rates, reserve ratios, and liquidity injections without staff intervention. Early trials at the Bank of England and the European Central Bank (ECB) indicate that AI can outperform human committees in crisis-response speed and accuracy. The next declared ambition is full autonomy, with AI ‘governors’ acting faster than markets can react.

Nevertheless, as of May 2026 today’s AI still cannot reliably automate 90 % of central-bank monetary-policy decisions in real time. Authoritative assessments from the Bank of England’s 2023–24 ‘AI Decision Lab’ project confirm that central banks continue to rely on human judgment to interpret simulation outputs, weigh risks that elude quantification, and ensure alignment with statutory mandates. Current AI models remain vulnerable to regime shifts, data gaps, and adversarial shocks—flaws that materially weaken claims for full automation. Policy experiments, such as the Bank of England’s lab, point toward high-value advisory roles but explicitly stop short of granting AI full autonomy.