What Google’s ‘century bonds’ say about AI infrastructure buildout

It is a striking paradox that Alphabet, a company built on the ephemeral shifting sands of software code, has chosen to issue financial instruments designed to outlast the vast majority of its employees. The tech giant’s decision to issue “century bonds”—debt that will not mature until the 22nd century—while sitting on a fortress balance sheet seems, at first blush, an act of financial eccentricity.

However, a closer inspection reveals a sophisticated exercise in capital structure optimization, driven by the unprecedented demands of the artificial intelligence arms race.

The logic of Alphabet is rooted in the sheer, eye-watering scale of the current capital expenditure cycle. With the company projecting up to $185bn in spending this year alone—a figure rivaled only by Amazon’s $200bn target—the cost of building the physical backbone of the AI era is beginning to resemble the GDP of a mid-sized nation.

While Alphabet possesses a formidable cash pile, draining it to purchase AI chips and construct data centres seems counterintuitive. By tapping the debt markets, the company is likely preserving its “dry powder” for strategic acquisitions and stock buybacks, essentially letting the bond market subsidise the construction of its AI infrastructure.

That the market is willing to lend to a technology firm for a hundred years is a testament to Alphabet’s transition in the AI era—from a faltering Bard launch to nearly leading the pack in AI tools. The resounding demand for the bond offering, which allowed the company to increase the raise to $20bn, suggests investors view Google’s search monopoly as a fixture as permanent as a tax base.

The company’s move to issue debt in British pounds and Swiss francs further underscores this ambition. It is a textbook “natural hedge”: by matching debt liabilities with the local currencies where they are pouring concrete for European data centres, Alphabet insulates itself from foreign exchange volatility while shopping for the lowest global yields.

But corporate history reveals companies that mistook momentary dominance for eternal relevance. In this aspect Motorola and JCPenney are cautionary tales. In 1997, Motorola was the undisputed king of mobile telephony when it issued its own century bonds; two decades later, it had been carved up and sold off, a victim of the smartphone revolution it failed to foresee. JCPenney’s bankruptcy occurred a mere 23 years into its 100-year bond.

The risk for Alphabet is that the silicon-based architectures needed for today’s Large Language Models may become obsolete in future—perhaps quantum or biological computing will lead the next phase of AI development—leaving the company’s data centres go out of use by 2050.

To understand how Alphabet views this risk, one might employ a Monte Carlo simulation, forecasting the “critical path” of the next thirty years. In the “Sovereign Tech” scenario—arguably the top quartile of outcomes—AI evolves into a high-margin utility comparable to electricity or the internet itself. Here, Alphabet’s $185bn annual bet creates an insurmountable moat, and the century bond becomes the “gold standard” of corporate debt, serviced effortlessly by the cash flows of an AI-driven economy. In this world, locking in today’s cost of capital for ten decades looks like a masterstroke.

However, the median outcome is more prosaic: a “Utility Squeeze.” In this scenario, the magic of AI fades into a commoditised necessity. Margins compress as hyperscalers bludgeon one another on price, and while Alphabet remains solvent, it morphs into a boring, regulated utility. The bond performs safely, but the company’s explosive growth days are over.

Lastly, the true danger lies in the “Obsolescence Trap”— the bottom quartile of probabilities where technological disruption renders current infrastructure worthless. Here, the century bond would trade at a deep discount as investors realise they funded the wrong revolution.

The century bond serves as a hedge against precisely this uncertainty. If the AI data centres bet pays off—very likely in the near-term—the debt will be negligible. If the technology stagnates or shifts, the bond provides a permanent capital base for a century-long runway that does not require refinancing during a potential liquidity crisis.

Alphabet is betting that by securing the money to build the future today, it can buy enough time to ensure it does not become the Motorola of tomorrow.

Published – February 19, 2026 09:01 pm IST