What if the most valuable AI in finance wasn’t the one making decisions but the one making you stop and think? That is the premise behind HI, or Human Intelligence, a prototype AI assistant currently under development by The Ethics Centre in partnership with Miroma Project Factory. Unlike the algorithmic engines that dominate trading floors or automate compliance checks, HI is designed to integrate ethical reflection directly into the fast-moving workflows of banking and finance professionals. It does not replace judgment; it strengthens it.
Its architecture is based on deep natural language processing, drawing on the context of live interactions via email, chat, or browser-based interfaces. Delivered as a lightweight widget and embedded into Slack or Microsoft Teams, HI pops up in real time with context-aware prompts at the point of decision-making-whether it’s a questionable client request, an interpretation of policy, or a product design choice. These are not prescriptive nudges but spur the user toward deeper consideration, drawing on principles from behavioral science similar to those deployed in gamified learning environments that reinforce critical thinking and collaboration.
The design of HI reflects a growing recognition in the financial sector that AI’s role is not just about operational efficiency. A 2024 Deloitte survey of Generation Z and millennials showed that half of Generation Z and 43% of millennials have rejected assignments on ethical grounds, with many having left employers entirely over values misalignment. Meanwhile, 77% of frontline finance staff report facing ethical dilemmas multiple times a day. The creators of HI saw an opportunity to embed decision support precisely at the moment of need, rather than relegating ethics to annual training modules or crisis meetings.
The most ambitious part of the system is its learning layer: interactive scenarios, micro-lessons, and decision-tree guidance in a gamification framework that rewards streaks of engagement and skill progression. That approach mirrors findings from educational research indicating such gamification methods combined with AI can significantly boost motivation, creativity, and problem-solving ability. Allowing users to work their way through ethically complex situations in a safe environment, HI lets users explore consequences without real-world risk, gradually building what its designers call “ethical muscle.”
From a technical governance standpoint, HI also aligns with the ethical AI frameworks now coming online in finance. Industry analyses have underlined the need for responsible AI in banking to be transparent, fair, and accountable: all principles which HI operationalizes through explainable prompts and user-controlled reflection logs. Unlike most existing uses of AI, let alone those deployed in high-consequence decisions about credit scoring or fraud detection, the design of HI avoids the “black box.” Working with CSIRO, The Ethics Centre is also taking care to ensure that the development of HI will meet regulators’ expectations-which are themselves still evolving, from the European Union’s risk-based classification and regulation of AI to sectoral guidelines issued within the United States and throughout the Asia-Pacific.
Indeed, results from beta testing of the prototype show strong traction: 95% of the users reported a better understanding of how their values drive decisions, 82% saw potential across the industry, and 91% rated the dashboard and feedback features positively. These metrics are notable, given the sector’s traditional skepticism toward “soft” skills training, and underscore a shift toward valuing ethics as a competitive advantage.
HI’s deployment strategy also addresses one of the key friction points to adoption: workflow friction. By meeting users in the platforms they already inhabit, it avoids the drop-off common to standalone training portals. Underlying NLP models are tuned for brevity and relevance so that prompts feel like a natural part of the workstream, not an interruption. It is critical in environments where decision latency can have financial consequences.
The big picture in all this is that AI in finance has grown rapidly, with Statista estimating sector spending to increase from USD 35 billion in 2023 to USD 97 billion by 2027, along with significant operational gains but also increased ethical risk. Algorithmic bias, data privacy breaches, misuse for market manipulations-these are not hypothetical but very real. Tools like HI reflect an alternative path for AI adoption: one where ethics are treated not as compliance overhead but as embedded infrastructure.
As Kat Robinson, Global CEO of Miroma Project Factory, said, This generation doesn’t just want to do good work – they want to do work that is good. Projects like HI are our answer to that challenge: a quiet, consistent companion that helps people think clearly, act thoughtfully, and lead with integrity in the flow of their day.
In a field where milliseconds can move markets, HI is betting that taking a few of those milliseconds back-for reflection-could reshape not just individual choices but the culture of finance itself.
