Yet behind the promise of innovation lies a deeper question, can machine-made decisions truly reflect the values we hold dear? This is not merely a matter of efficiency, but a pursuit of virtue in a digital era.

This question calls us back to conscience. Aristotle taught that virtue is not granted overnight. It is the result of habit and a conscious will to do what is right. For him, virtue lies in balance, in the golden mean between extremes.

In the context of AI in banking, that balance means finding harmony between technological efficiency and social justice. Between innovation and protection for the vulnerable. Without policies that ethically guide AI, we risk building systems that are intelligent, but lacking in wisdom.

Immanuel Kant, the 18th-century German philosopher, reminds us that moral action must stem from a rational duty to uphold human dignity. His principle that people must always be treated as ends, not means, rings especially true.

Customers are not just data points in an algorithm. They have the right to understand how decisions are made about them, to receive explanations and to raise objections when harmed. In an AI-powered world, this means building systems that are transparent and accountable.

The Financial Services Authority (OJK) has encouraged accessible complaint mechanisms. What’s needed now are simpler, app-based channels that make complaints easier. Transparency must meet people where they are. Not just in regulations, but in everyday experience.

René Descartes once said that doubt is the first step toward clarity. In the face of innovation, we must ask: are AI decisions truly better, or simply faster and cheaper? This kind of skepticism is not a hurdle, but a safeguard, a way to keep technology on the right track.

On April 29, the OJK issued the Banking AI Governance Guidelines, offering ethical and practical direction for the use of AI in banking. More than rules, the guidelines remind us that AI is not just a tool. It is a mirror of the values we choose to uphold.

The data used to “train” AI is, in itself, a reflection of the past. It holds the patterns, judgments and biases of systems that came before. If that data is skewed, so too will be the decisions AI makes. Perhaps even amplifying existing bias.

Take credit scoring algorithms. If they rely solely on formal financial history, informal workers in rural areas could be excluded. Without inclusive approaches, AI risks deepening inequality rather than narrowing it.

Aristotle’s notion of virtue as balance reminds us. AI in banking must avoid extreme profit obsession and operational rigidity. It should foster human flourishing by offering equal access to financial services. Whether for an ojek driver in Jakarta or a rice farmer in Nusa Tenggara.

Transaction data must not be used to exclude. Instead, it should help us understand and empower. It should uplift micro-businesses in remote villages, not leave them behind. Without oversight, these systems may favor the profitable few, quietly sidelining the rest.

Kant’s call to “treat people as ends” compels clarity in automated decisions. When an AI system rejects a loan or flags a transaction as suspicious, customers deserve an explanation. Without clarity, they are reduced from active participants in a financial system to faceless data points.

Jean-Paul Sartre once said that “man is condemned to be free”. In his view, there is no divine script or fixed nature to fall back on, only the task of choosing. With it comes the burden of becoming who we are through those choices. But what happens when decisions that shape our lives are no longer ours to make?

When an algorithm denies a farmer credit because his income is seasonal, or flags a street vendor’s transaction as suspicious without context, freedom becomes something distant. In Indonesia, where digital literacy remains uneven, many are not even aware that their financial fate may rest in the logic of a machine. At that point, we are not choosing, but being chosen for.

Descartes would tell us to begin with doubt. Not to reject technology but to question it, so we do not lose our way. Just last month, OJK and Bank Indonesia warned of fake AI-generated receipts used by scammers. The same tools that promise progress can also deceive. Audits, then, are not just procedures. They are a form of moral attention.

Indonesia’s challenges in adopting AI are uniquely layered. Beyond limited digital literacy, many customers still lack the power to question or negotiate. In this setting, trust is not a luxury, but a moral necessity. Ethics is not a technical supplement, but the very foundation on which a just financial system must be built.

This is not a clash between humans and machines. It is about virtue, embedding empathy, fairness and integrity into every line of code and policy.

Indonesia is vast, diverse and unequal, from bustling cities to remote villages. The ethical challenges of AI are magnified by this landscape. Digital justice is not a lofty ideal. It is a pressing need for millions who still live outside the formal financial system.

The future of banking cannot be left to algorithms operating in the shadows. We need collective action. The government must strengthen oversight. Banks must invest in systems that are open and inclusive. The public must be aware of their rights.

In an era dominated by smart machines, we must allow our conscience to guide us. Aristotle reminds us that virtue is a practice in balance. Kant calls us to honor human dignity. Sartre reminds us about freedom to choose. Descartes urges us to question, so that clarity may follow.

If we weave these lessons together and root them in the Indonesian reality, we can build a financial system that is both sophisticated and humane. A system that is just, not just efficient.

---

The writer is a legal and corporate secretary at Bank DBS Indonesia. The views expressed are personal.