How AI is Transforming Infrastructure Finance

Why infrastructure finance is finally ready for its AI moment—and what that means for getting real projects off the ground faster, smarter, and with greater impact.

There are moments in your professional life when two very different threads of experience converge—and the result feels inevitable. That’s how it’s been for me and Stanley Boots. After years spent on opposite ends of the infrastructure finance value chain—he in legal and regulatory reform, me in engineering and advisory—we found ourselves sitting side by side again, this time as co-founders of Silta Finance. Our shared frustration with the inefficiencies in infrastructure delivery had finally met its match: artificial intelligence.

We hosted a LinkedIn Live recently to dig into this, not because AI in infrastructure is a fashionable topic—it is—but because it’s becoming a necessary one. And while most headlines focus on generative AI in creative industries, or automation in white-collar jobs, the transformation happening in infrastructure finance is arguably more consequential. We’re talking about a sector that underpins everything from access to clean water and public transport to climate resilience and basic economic opportunity. And yet, it’s a sector that still leans heavily on PDFs, manual checklists, and years-long timelines just to get a project through pre-feasibility. It’s a laggard. But it won’t stay that way much longer.

Stan and I have seen this evolution from the ground up. He started his career as an international lawyer working with Southeast Asian governments to draft the frameworks that make public-private partnerships (PPPs) viable in emerging markets. He’s worked on everything from feasibility requirements to contract structuring in Vietnam, Cambodia, Laos, and Myanmar. I, on the other hand, came into infrastructure through engineering. My early career saw me working for major consultancies—Halcrow, Atkins, EC Harris—producing feasibility studies, acting as a transaction advisor, and guiding lenders through due diligence for large-scale infrastructure across Europe and Asia. Eventually, I gravitated toward sustainability projects—coastal ports, fisheries, ocean infrastructure—because that’s where I felt the work really mattered.

It was on a driving range in Hanoi, sometime around 2012, where we first started asking ourselves: isn’t there a more efficient way to do all this? Why does it take so many months, sometimes years, and so much money to get even a single infrastructure project to bankability? Why are so many well-intentioned ideas buried under the weight of their own procurement processes?

Those conversations planted the seeds for what would eventually become Silta. Years later, after my own journey through blockchain and tokenization, and now with the rise of practical AI tools, we’re back together, building systems that help solve those same frustrations.

AI, applied properly, is finally offering a serious answer to the inefficiencies we used to fight manually. And it’s starting to show its value in four critical areas of infrastructure finance: tendering, ESG and safeguard assessments, early-stage project development, and stakeholder communication.

Take tendering, for example. It’s one of the most painful parts of any infrastructure project, whether you’re the one submitting a bid or the one evaluating it. The process is incredibly document-heavy. Thousands of data points. Every sentence in every document is a potential compliance issue. Every CV has to meet specific requirements. And then there’s the maddening time pressure—tight deadlines, misaligned teams, and high stakes. I still remember working on the LRT Line 1 bid in the Philippines, coordinating between two massive conglomerates, each with internal silos. We spent days just trying to locate basic data—project precedents, land ownership details, even advertising rights around train stations. We finally got the last bit of critical information late on a Friday night, and spent the entire weekend restructuring financial models. The bid went in just under the wire. But it was madness.

With AI, that same bid could have been built in days. By connecting corporate data rooms to an AI-powered interface, we could’ve surfaced relevant precedents instantly. We could’ve evaluated compliance with bid requirements automatically. We could’ve scored ourselves against evaluation criteria, and, had the government’s transaction advisor challenged any of it, we’d have had a full audit trail ready to go. That’s the kind of transformation AI enables—not magic, not science fiction, just better systems.

The same applies to the government side. Drafting bid documents, managing deviations from standardized contracts, checking CV compliance, managing conditions precedent—it’s all tedious, high-stakes work. And too often, public authorities lack the resources to do it well. AI can change that. In fact, with a decent set of background documents and goals, even a non-technical civil servant could generate a robust first draft of bid rules. From there, the AI can check for consistency across the entire procurement package and flag any anomalies. That means stronger procurement, faster evaluations, and less risk of delays due to non-compliance or legal uncertainty.

And then there’s ESG. Ten years ago, ESG reporting was mostly a nice-to-have, occasionally funded out of a company’s marketing budget. It certainly wasn’t taken seriously by regulators. But that’s changed. Today, countries like Japan are introducing mandatory green transformation (GX) reporting. Multilaterals are demanding alignment with the IFC Performance Standards. And even consumers are getting more vocal about sustainability. That’s all good news—but it also creates complexity. How do you measure environmental and social risks consistently across thousands of infrastructure projects? How do you track compliance over years? How do you go beyond risk to measure “additionality”—the positive impacts, like access to clean water or local job creation?

Enter Silta Finance

That’s the work we’ve done at Silta. We reverse-engineered the IFC standards into AI-readable rating models. This lets us evaluate ESG and safeguard performance based on uploaded project documents, flag gaps, and track improvements over time. No AI can replace local knowledge, of course. It can’t walk around a site. But it can assess what it’s given—and it gives the human assessor more time to do the things that matter: the interviews, the observations, the community consultations. It’s a better division of labor.

This matters even more when you consider early-stage project development. In many emerging markets, the biggest bottleneck isn’t investor interest—it’s that local governments don’t have the capacity to put forward well-structured, fundable projects. Take General Santos, a city in the southern Philippines. They’ve wanted to run PPPs for years, but didn’t have the staff or resources to get past square one. Now, using platforms like ours, even a small team can draft business cases and pre-feasibility studies with AI support. We’ve tested this internally. Feed the AI a few details—say, a 10MW solar project in Palawan—and it can produce a 25-page draft report, complete with comparable projects, regulatory overviews, and potential contractors. Is it a finished PRE-FS? No. But it’s a massive head start. It makes the impossible—suddenly possible.

And that’s where AI’s biggest impact might lie: not in replacing jobs, but in closing the capacity gap. At the UN event I recently attended, we discussed the global shortfall in infrastructure investment. It’s around $1 trillion per year. There simply aren’t enough consultants in the world to hit that target with business-as-usual methods. Even the multilaterals that have committed to $100 billion in green financing by 2030 are nowhere near that throughput today. They close one or two deals a year, not thousands. Unless someone starts writing $20 billion checks per deal, it won’t scale. But with AI, we can raise our throughput without raising burnout.

Which brings me to a final point: the human side of this. There’s a natural fear around AI. It’s easy to believe it’s coming for your job. But in infrastructure finance, it’s not replacing people—it’s relieving congestion. It’s removing the bottlenecks that stop us from delivering the clean energy, water access, waste systems, and transport infrastructure the world needs. It’s taking the 14-hour days and replacing them with something more humane. It’s letting consultants go home, see their families, maybe even play football in the evening. Or, if they choose, take on more projects and expand their impact.

The legal world is already grappling with this: should AI mean lawyers do 10,000 hours of work per year for the same 1,700 hours billed? Or should we shift the billing model entirely and just let lawyers live better lives while still delivering excellent value? It’s a live debate. But for consultants in PPPs, the answer feels obvious. We need more throughput. We need better tools. And we need a generation of public and private actors who are equipped, not overwhelmed.

That’s the vision we’re building toward at Silta. Not a future where AI runs everything. But one where people like us—consultants, engineers, civil servants, entrepreneurs—can finally move faster, think clearer, and get infrastructure on the ground where it’s needed most.

Not tomorrow. But now.

If you’re working in infrastructure finance—whether in government, consulting, or investment—and you’re grappling with the realities of capacity, delivery, or how to practically deploy AI in your workflow, I’d love to connect.

Drop me a note at ben@silta.finance. Let’s explore how we can accelerate your projects, together.