The Future of Finance Is Autonomous and Always-On

Why finance is no longer a machine, and leadership can no longer pretend otherwise

Five years ago, the world operated at a fundamentally different tempo. Not slower in absolute terms, but slow enough for human intuition to remain useful. Financial systems moved at a pace where leaders could still rely on experience, pattern recognition, and post-hoc oversight. We were in the first half of the chessboard: linear progress, predictable feedback loops, time to react, interpret, and course-correct.

That world is gone.

Today, the convergence of technologies has pushed us decisively into the second half of the chessboard. Change is no longer incremental; it is compounding. Systems evolve faster than organisations, and the humans inside them, can cognitively absorb. This is not simply a story about speed. I’ve always insisted that speed is the wrong obsession. What truly matters now is autonomy.

Financial systems are no longer merely executing predefined rules. They are beginning to decide, adapt, and optimise on their own.

I first experienced what a living system feels like far from finance. In 2011, I cycled more than 14,000 kilometres around Australia in 100 days. The journey was not reckless. It was meticulously planned: routes, energy expenditure, recovery, contingencies, everything was thought through. And the planning worked, until it didn't, because the harsh Australian outback is unpredictable and conditions are changing all the time.

What stayed with me wasn’t resilience or endurance. It was cadence and adaptability. Each day required discipline, constant sensing, small adjustments, and the humility to respond to conditions as they unfolded. You don’t control a living system by setting a plan and walking away. You stay inside the loop.

That is exactly the shift finance is undergoing now.

The Living Financial System

For decades, financial markets were governed as static systems designed for periodic oversight: batch settlement, delayed reconciliation, committee-based governance. Today, they are becoming continuous, adaptive systems that move whether we are ready or not. The uncomfortable truth is that we did not consciously enter the digital age,  we sleepwalked into it. Governance models, incentive structures, and career paths evolved incrementally, while technology compounded exponentially.

As data became the new oil, relevance shifted from access to interpretation. Now, as machines enter the decision-making loop, interpretation itself is being compressed. Decision cycles that once took days now take seconds. Human oversight increasingly happens after execution, not before it. The more decision time collapses, the more important accountability becomes, and yesterday’s systems were never designed to deliver accountability at machine speed.

This raises the defining question for today’s leaders: when machines decide faster than humans can reflect, what is leadership actually responsible for?

From Financial Infrastructure to a Computable Economy

We are witnessing the rapid maturation of multiple technologies simultaneously. Each one does not merely add change; each breaks a foundational assumption that once underpinned financial stability. Together, they are forming a computable economy: a system that operates continuously, automatically, and at machine speed.

Tokenisation removes friction. Settlement and reconciliation, once measured in days, now collapse into seconds. The market for tokenised assets has already grown from roughly $860 million in 2023 to more than $2.3 billion by mid-2025, and that trajectory is accelerating. Major institutions are no longer experimenting at the edges. At the same time, tokenized real-world assets have reached $33 billion in value as of October 2025, enabling micro-transactions, global investor access, and reduced minimum investments (from millions to hundreds of dollars).

BNY Mellon has announced real-time, on-chain deposit settlement capabilities. Nasdaq plans 24/5 trading with continuous clearing by the end of 2026, and the New York Stock Exchange’s new digital platform will enable 24/7 operations, instant settlement, orders sized in dollar amounts, and stablecoin-based funding. NYSE is building a new way to bring equities on-chain AND the venue to trade them.

At Davos 2026, stablecoins were treated less like “crypto” and more like a monetary technology, as an IMF panel warned they can pressure weak fiscal/monetary regimes and potentially pull deposits from banks in emerging markets. The framing at the World Economic Forum was pragmatic: banks bring trust, compliance and risk controls; blockchains bring programmability and always-on settlement.

Artificial intelligence removes interpretation. Decision-making shifts from explicit human logic to probabilistic machine judgment. AI is no longer confined to back-office optimisation; it is moving to the front lines of finance. Chatbots like Bank of America’s Erica and HDFC Bank’s EVA already handle millions of customer interactions each month, orchestrating service, risk, and engagement at scale. This is not about cost reduction alone. It is about redefining how decisions are made and executed.

Quantum computing threatens the assumption that privacy and encryption endure. The belief that cryptography provides long-term security is becoming fragile. Even permissioned blockchains must now be designed with quantum-resistance in mind. Security is no longer a static property; it is a moving target.

In this emerging system, everything becomes interconnected. Multiple parties share a single version of reality. Process fades into the background. What remains is responsibility and accountability,  enforced at the speed of light.

And that leads to a question most leadership teams are not yet prepared to answer: if the systems we are building are faster, cheaper, more autonomous, and increasingly opaque, what exactly are we training future leaders to do?

When Risks Converge, Systems Break Differently

Each of these technologies introduces material risk on its own. The real danger emerges when those risks collide. We are no longer managing technological risk in isolation; we are confronting systemic risk born from convergence.

Automation, AI, and robotics are hollowing out traditional career ladders — especially the entry-level roles that once trained judgment through repetition. European banks alone are planning to cut more than 200,000 jobs by 2030 as AI takes hold, with back-office, risk management, and compliance functions seeing efficiency gains of up to 30%. ABN AMRO plans to reduce its workforce by a fifth by 2028. Société Générale’s CEO has been blunt: “Nothing is sacred.”

McKinsey estimates that roughly 30% of U.S. jobs are automatable. Goldman Sachs projects that up to 50% of work tasks could be fully automated by 2045. This is not just a labour story; it is a leadership story. When entry-level roles disappear, organisations lose the apprenticeship pathways through which judgment is formed.

At the same time, money itself is beginning to privatise. Stablecoins and tokenised liquidity are no longer fringe experiments. The stablecoin market already exceeds $250 billion and is projected to reach $2 trillion by 2028. In 2025 alone, more than 19  stablecoin payment players raised over $1.5 billion in funding. The passage of the GENIUS Act in July 2025 marked the first major U.S. regulatory framework enabling bank-issued stablecoins. Monetary assumptions that held for decades are now up for renegotiation.

Cyber risk is evolving just as fast. AI-driven attacks operate at machine speed, eroding trust in data itself. Small criminal teams equipped with AI tooling can now operate at the scale of much larger organisations, automating fraud with astonishing realism. This is already recognised as one of the most disruptive forces in financial cybersecurity.

All of this converges faster than governance frameworks can adapt. Around 70% of banking firms experimenting with agentic AI still lack robust governance structures. 2026 is the moment where political, financial, legal, operational, and technological stresses surface simultaneously and reinforce one another.

Oversight remains episodic while systems learn continuously. For the first time, the systems we oversee are adapting faster than the humans responsible for them can cognitively absorb.

Designing the Financial System We Will Inhabit

We are now in an unstable transition phase. The old system is eroding, and the new system has not yet been consciously designed. This is where unintended consequences emerge, not from malice, but from misaligned incentives.

In a financial world still dominated by short-term performance metrics and quarterly pressures, the challenge is not technological capability. It is intent. A living financial system must be trustworthy, inclusive, and resilient by design.

By 2035, financial markets will operate continuously, across borders, twenty-four hours a day. The question is not whether this happens, but how. Governance cannot remain something bolted on through policy documents and committees. It must be embedded into system architecture itself. Accountability must be designed in before automation, not retrofitted after systems are already in motion.

This is the true leadership challenge of the coming decade. The question is no longer what technology makes possible. It is what kind of system we are willing to legitimise.

The Choice We Can No Longer Avoid

In a world of converging systems, the trusted convener, the institution or leader who understands the living system as a whole, becomes indispensable. Tokenisation, AI-driven markets, and quantum computing are not distant futures. They are already here, quietly reshaping the foundations of finance.

The real choice is design versus drift.

Will we deliberately build systems aligned with societal benefit, accountability, and long-term resilience? Or will we inherit systems shaped by inertia, short-term incentives, and machines optimising for objectives we never fully defined, nor fully understood?

In adaptive systems, relevance no longer comes from having the answers. I continue to argue that it comes from deciding which questions the system is allowed to ask.

And that decision cannot be automated.