By Anil Wasif

May 13, 2026

PARIS — The intelligent design of boulevards that makes this among the most meticulously planned cities on Earth was produced by a man who understood that infrastructure is how a society foresees its future.

Baron Georges-Eugène Haussmann was appointed prefect of the Seine in 1853 with the mandate to remake Paris. He demolished thousands of buildings and imposed a building code that dictated facade materials, cornice heights, balcony placement, and roof angles across thousands of properties.

The result: limestone facades, iron balconies on the second and fifth floors, zinc mansard roofs, and a uniform skyline that holds the city together from any vantage point.

A century later, a young man named Yves Saint Laurent arrived from Oran, Algeria, became chief designer at Dior at twenty-one, then started over with his own name. He put women in tuxedos, called it Le Smoking, and proved that structure could liberate rather than constrain. Haussmann designed the boulevards. Saint Laurent designed how people walked down them.

Both built systems, not objects, and the love in both was structural: in the cornice lines that give Paris its horizon and the cut of a jacket that has not gone out of season since 1966.

I spent the first week of May here for the OECD’s annual infrastructure week, my third year at the table. I serve on two of its working bodies: the Network of Senior Infrastructure and PPP Officials, the SIP, which shapes infrastructure governance across the membership, and the Infrastructure Delivery Professionals network, the IPD, which convenes practitioners from procurement and delivery agencies.

The SIP debates direction. The IPD tests it against reality. The proceedings are confidential, but I can report on my favourite public session of the week: the AI in Infrastructure Resilience panel.

The OECD is revising its Recommendation on the Governance of Public-Private Partnerships for the first time in a decade.

Private investment in core infrastructure has trended downward for nearly 20 years. Fewer firms are bidding. The traditional model assumed risks could be identified and priced at contract signature. The evidence says otherwise. The OECD’s STEPS procurement strategy framework points toward value for money managed continuously across a project’s life, with commercial terms that evolve as unknowns become known.

Later this month, the senior budget officials meet ahead of the June 3-4 OECD Ministerial Council Meeting under the theme Restoring Public Finance. The Senior Infrastructure and PPP Officials (SIP)’s value rests on something no country or a think tank can produce alone: comparative evidence about which delivery models are working under which conditions and why.

The AI panel began with an admission that carried more force than any procurement argument. Infrastructure is one of the least digitally mature sectors in the global economy. The technologies capable of transforming how we plan, build, and maintain physical systems are proven and producing results. They are also, almost everywhere, sitting on the shelf.

Singapore has built Virtual Singapore, a national-scale digital twin integrating urban planning, transport modelling, water management, and climate resilience into a single platform. Norway has deployed digital road maintenance. South Korea’s K-water is rolling AI-managed water treatment across its network, with a certification system launching this year. The presenters were reporting from programs that work.

The argument that emerged aligns with Haussmann’s approach to Paris: a shift from managing individual assets to managing systems. Digital twins and machine learning make it possible to see infrastructure as a whole; where failures cascade, where demand shifts with weather and demographics, where the maintenance decision made today determines the resilience available in a decade.

Haussmann’s building code did the same thing in the vocabulary available to him. Thousands of private property owners, each with their own architect, producing buildings that worked together because the framework was designed at the system level.

Digital twins and machine learning make it possible to see infrastructure as a whole; where failures cascade, where demand shifts with weather and demographics, where the maintenance decision made today determines the resilience available in a decade.

The boulevards were ventilation corridors for a city suffocating in medieval density. The parks were public health infrastructure. The uniform cornice line was the interface between private construction and public coherence.

What Singapore and South Korea are building now is the digital expression of the same conviction: design the system well, and the assets within it become capable of serving people whose needs cannot be anticipated at construction.

Saint Laurent understood this in his own medium. He designed not garments but a vocabulary of shapes and proportions that other designers could extend long after he put down his pencil. The Mondrian dress of 1965 was a proof-of-concept for an idea about simplicity that is still generating new work 60 years later.

The panel also surfaced what should be a policy priority for every OECD government. The investment case for digital infrastructure is not speculative. The adaptation-and-resilience economy is growing, and the countries that move first on industrial strategies for infrastructure technology will define the market for everyone else.

Regulators could be rewarding utilities for reducing system risk and improving climate adaptation instead of measuring compliance against static benchmarks. Outcome-based regulatory models, where performance against resilience targets determines the return, would redirect capital toward the technologies that are already proven.

The UK energy regulator is consulting on exactly this for its distribution companies. The question is not whether the technology works. It is whether the policy environment will let it.

The barriers to digital adoption are institutional, financial, and regulatory. Procurement frameworks reward lowest initial price. Returns on digital infrastructure accrue across the long life of an asset, invisible to systems designed around short-term cost. No single country has solved these problems alone. Norway learned from the UK. South Korea benchmarked against Singapore.

The IPD’s multi-year permitting research and its work on public buyer maturity produce operational knowledge rather than communiqués. This is multilateral cooperation as it is supposed to function: practitioners sharing what works and what failed, in a setting where candour is protected.

When budgets tighten, infrastructure is often the first line that gets cut because the consequences take years to become visible. A deferred maintenance decision does not show up in the next quarter. A country that fails to digitize its asset management will not know what it lost until something fails that should not have failed.

This is the structural challenge facing every G7 government right now.

Central government debt across the OECD reached $17 trillion in 2025. Bond markets are paying attention. The political incentive is to spend less on things voters cannot see. And infrastructure, by definition, is the thing that works so well that nobody notices it until it stops.

The Canadian government under Mark Carney has prioritized infrastructure and major projects as a matter of sovereignty. It’s innovative framing for a policy portfolio involving bricks, mortar, tracks, pipelines, minerals and other very prosaic systemic assets.

That innovation should extend to how all the human capital invested is mapped, designed, measured and maintained.

Policy Columnist Anil Wasif is a public servant in the Ontario government. He serves on the University of Toronto’s Governing Council and the Advisory Board of McGill’s Max Bell School. Internationally, he serves on the OECD’s Infrastructure Delivery Committee and the Board of Trustees at BacharLorai Global. The views expressed are his own.