Central bankers eye higher rates as AI boom fuels scramble for cash

The AI boom is acting like a major economic demand shock. Governments and companies are spending heavily on chips, data centres, power networks and telecommunications before AI has produced its promised productivity gains. This raises demand for workers, materials, energy and financing now, while the supply benefits may come later.

That timing matters for central banks. Federal Reserve Chair Kevin Warsh said investment was hitting the demand side before the supply side. When many borrowers compete for limited capital, lenders can charge more. Higher spending can also add pressure to prices, making inflation harder to control.

Central bankers therefore see a possible shift away from the cheap-money era after 2008. The neutral interest rate, or R-star, may be higher than previously believed. If so, rates that once seemed restrictive may no longer cool demand enough. The Bank of Japan has already cited AI-related demand when raising rates, while markets expect further moves from several major central banks.

The phrase describes intense competition for financing around the world. Governments are borrowing for defence, industrial policy and energy projects, while companies are funding AI-related infrastructure. Both groups want money at the same time, increasing pressure on lenders and investors.

The article identifies AI investment as a key driver. Projects involving silicon chips, data centres, power grids and telecoms networks require large upfront spending. Investment is concentrated especially in the United States, with Japan and other parts of Asia also involved. More borrowers chasing capital can raise interest rates and bond yields.

This is the reverse of the post-2008 concern about excess global savings and inflation that was too low. ECB policymaker Martins Kazaks said there is now a global fight for money. The shift matters because governments already carry large debts, and more expensive financing can squeeze public budgets while central banks try to contain demand and inflation.

The AI-related investment wave is broad rather than limited to software companies. Governments and businesses are funding silicon chips, data centres, power grids and telecommunications networks. These projects support AI directly or indirectly, so the spending reaches manufacturing, energy, infrastructure and communications.

Its geographic scale is global, but the United States is the main centre of the surge. Japan and other parts of Asia are also receiving significant investment. In Europe, ASML and Siemens are identified as among the few clear winners. The article does not provide a total dollar value for the investment wave.

This breadth helps explain why central bankers treat AI as a macroeconomic shock. Spending is not confined to one technology sector. It increases demand for capital across several industries and can influence energy use, industrial activity and financial markets. Because investment comes before productivity gains, the near-term economic effects may be stronger than the later benefits.

Heavy investment increases demand immediately. Companies order equipment, build facilities and compete for workers, energy and materials. Governments also borrow for defence, industrial policy and energy projects. These demands can strain available resources and push prices higher before AI improves production capacity.

The key mechanism is timing. Federal Reserve Chair Kevin Warsh said the AI boom was hitting the demand side before the supply side. Investors and borrowers therefore seek funds now, while the productivity gains promised by AI remain uncertain and delayed. Competition for capital can lift interest rates and bond yields, adding to project costs.

Central bankers view this as a fresh inflation risk. The Bank of Japan said investment-driven price pressures had become broader and more persistent. If inflation stays firm, policymakers may need to maintain higher rates. That could make the transition bumpy, especially if investors reassess optimistic AI earnings expectations and financial markets sell off.

R-star is the estimated neutral interest rate. It is the level that keeps an economy on an even keel, neither adding too much stimulus nor applying too much restraint. Central bankers use it as a guide when judging whether current rates are appropriate.

AI investment and government borrowing may increase demand for capital. That can raise the economy’s neutral rate. If the neutral rate is higher, a policy rate that once seemed restrictive may actually be closer to neutral. The same numerical interest rate would then slow spending less than policymakers expect.

This makes rate-setting more uncertain. South African Reserve Bank Governor Lesetja Kganyago said R-star was higher, but its exact level was obscured by “thick clouds.” Bank of England Chief Economist Huw Pill said rates might average a little higher. Central banks may therefore need higher rates to control inflation without assuming their old benchmarks still apply.

Governments with large debts are vulnerable because rising interest rates increase the cost of borrowing and refinancing. More public money must go toward interest payments, leaving less room for services, investment or emergency support. The pressure grows when bond yields rise across many countries.

The article links this risk to years when interest rates stayed near zero. Governments built up large debts while financing was unusually cheap. Now sovereign bond yields have climbed, with some reaching levels not seen since before the global financial crisis. New borrowing becomes more expensive, and existing debt may also cost more as it is renewed.

The euro zone is seen as particularly exposed, especially debt-laden France. The bloc also faces higher energy costs from wars in Iran and Ukraine, which add inflation pressure and squeeze public finances. ECB President Christine Lagarde said France’s first solution lies at home through fiscal policies, meaning tighter budget management may be needed.

Interest rates are a central bank’s main way to influence economic demand. Higher rates make borrowing more expensive and can slow household spending, business investment and government financing. Lower rates generally make money cheaper and can support demand, investment and growth. Policymakers seek a balance that avoids both excessive inflation and unnecessary weakness.

The AI boom makes that task harder. Spending on chips, data centres, energy and telecoms is strengthening demand before productivity gains arrive. Central banks may need higher rates to prevent this demand from creating persistent inflation. Yet they must also avoid choking off productive investment that could improve future supply and long-term growth.

Officials are uncertain because the correct neutral rate may have risen. The Bank of Japan has already raised rates, while economists expect the Bank of England to move on November 5. The ECB and Federal Reserve raised rates in September, and markets expect another move from both in December.