The wallet that knows more than your bank
For centuries, creditworthiness has been an imperfect science. Lenders have relied on backward-looking indicators (income, collateral, repayment history) assembled into scores that attempt to predict future behaviour. These systems, whether embodied in the US FICO model or UK credit reference agencies, such as Experian, Equifax or TransUnion, are often blunt instruments: incomplete, static and geographically constrained. Now imagine a system that observes not only what you have done, but how you behave in real time. A system that captures every transaction, every counterparty interaction, every financial decision and uses this data to infer not merely outcomes, but intent. This is the premise of the “insured digital wallet” combined with a “wallet of intent”: a concept that sits at the intersection of regulated AI, decentralised finance and digital identity. It is both an evolution of existing trends and a potential inflection point in how credit, compliance and trust are assessed.
From data to judgement: the role of FinLLM
At the centre of this architecture lies a new generation of domain-specific artificial intelligence. General-purpose large language models (LLMs) have demonstrated impressive capabilities, but they struggle with the precision, accountability and auditability required in financial services. This is where systems such as FinLLM, developed in alignment with UK regulators including the Financial Conduct Authority (FCA) and Prudential Regulation Authority (PRA), begin to matter. FinLLM is trained exclusively on regulated financial data and designed to meet compliance standards from the outset. The distinction is subtle but important - rather than retrofitting compliance onto a general model, FinLLM embeds regulatory logic into its training data and architecture. This includes alignment with frameworks such as the EU AI Act and the UK’s Consumer Duty regime. In practice, this allows the model to interpret complex financial data, policy documents, transaction histories and onboarding forms whilst maintaining traceability and auditability. It is, in effect, an AI designed not simply to answer questions, but to justify its answers.
The emergence of on-chain credit
The second building block is already visible in decentralised finance. Platforms are increasingly experimenting with on-chain credit scoring, using blockchain transaction data to assess risk. For example, Cred Protocol analyses wallet activity to estimate the probability of default or liquidation, generating a credit score based on behavioural patterns. Similarly, Centic has developed a crypto credit score model that mirrors traditional FICO ranges, applying them to on-chain activity. These systems rely on the transparency of blockchain data. Transactions are immutable, timestamped and publicly verifiable, so allowing for a level of granularity that traditional credit files cannot match. Moreover, the market for such products is growing rapidly. According to Huma Finance, the tokenised private credit market has expanded significantly in recent years, reflecting demand for alternative lending models.
From behaviour to intent
However, these systems remain limited, focussing on outcomes (whether a borrower repaid a loan, whether a position was liquidated), rather than the underlying behaviour that led to those outcomes. The “wallet of intent” concept goes further - it seeks to interpret patterns of behaviour as signals of future reliability. Frequency of transactions, diversity of counterparties, participation in protocols, consistency over time, all become inputs into a dynamic assessment of creditworthiness. Hence, this is where AI models such as FinLLM become critical. Interpreting such data requires not only statistical analysis but contextual understanding: distinguishing between strategic activity and opportunistic behaviour, between risk-taking and recklessness. In this sense, the wallet becomes more than a ledger, it becomes a narrative - a continuous record of financial decision-making.
Insurance: the missing layer
The addition of insurance transforms this concept from theoretical to practical. In traditional finance, insured deposits underpin trust. In the UK, the Financial Services Compensation Scheme protects deposits up to £120,000. Cryptoassets, by contrast, have historically lacked such protection. This gap has been a barrier to mainstream adoption; however, the regulatory landscape is evolving. The UK’s Property (Digital Assets etc) Act 2025 recognises digital assets as a form of property whilst the FCA is developing a framework for cryptoasset regulation, expected to come into force by October 2027. These developments create the possibility of insured digital wallets - regulated custodial solutions that combine asset protection with compliance oversight. Private initiatives, such as Bitget’s $600+million protection fund, hint at how such models might operate.
Insurance introduces a crucial feedback loop - if risk can be measured, through a wallet’s behavioural and intent-based data, it can be priced. The wallet’s credit score becomes not merely a signal to lenders but an input for underwriters
The commercial opportunity
For financial institutions, the implications are significant:
- credit assessment could become more dynamic and inclusive - individuals without traditional credit histories, particularly in emerging markets, could build reputational capital through their wallet activity. This aligns with broader efforts to expand financial inclusion.
- compliance could become more granular - rather than relying on periodic checks and static documentation, regulators could monitor activity in near real time, supported by AI systems capable of flagging anomalies and risks.
- new revenue streams could emerge - wallet providers could offer credit scoring APIs, insurance products and data analytics services, creating a layered ecosystem around digital identity and financial behaviour.
The opportunity is not hypothetical - it reflects the convergence of technologies and regulatory frameworks that are already in motion.
The risks: privacy, gaming and power
The risks are equally profound:
- privacy - a system that records and analyses every transaction raises fundamental questions about data ownership and control. The tension with data protection frameworks such as the UK GDPR (particularly the “right to be forgotten”) is unresolved.
- gaming - just as individuals learn to optimise traditional credit scores, sophisticated users could structure transactions to create a favourable “wallet of intent”, without reflecting genuine behaviour.
- cross-chain fragmentation - activity on one blockchain does not automatically translate to another, complicating the creation of a unified credit profile.
- power - whoever controls the infrastructure (wallets, scoring models, insurance mechanisms) will wield significant influence over access to credit and financial services.
The cost of inaction
For institutions, the greatest risk may not be the technology itself, but the failure to engage with it. Banks that ignore on-chain credit systems may find themselves bypassed in lending markets, insurers that fail to develop models for digital assets may cede ground to new entrants, and regulators that delay may struggle to shape standards that are set elsewhere. The history of financial innovation suggests that early adoption confers lasting advantages. Payment networks, credit card systems and digital platforms have all demonstrated the power of network effects. Once a standard is established, it becomes difficult to displace - and the same may be true for wallet-based credit systems. If a particular model achieves scale (combining data, AI and insurance), it could become the default infrastructure for digital finance. The emergence of the “wallet of intent” raises a broader question: what kind of financial system do we wish to build? One possibility is a system that is more efficient, inclusive and transparent, where individuals control their data and benefit from their financial behaviour. Another is a system that concentrates power, where surveillance and control are embedded into the architecture of money itself. The outcome will depend on design choices: how data is stored, who has access, how models are governed and how risks are shared.
The next layer of finance
The insured digital wallet is not merely a product innovation. It represents a shift in how financial trust is constructed, moving from static records to dynamic, data-driven assessments of behaviour and intent. Combined with regulatory-aligned AI systems such as FinLLM, it has the potential to redefine credit, compliance and access to finance. But it also introduces new risks, new dependencies and new questions regarding power. For institutions and policymakers, the challenge is not simply to adopt or reject these technologies. It is to shape them, to ensure that the next layer of financial infrastructure reflects not only efficiency, but accountability and fairness.
Ultimately, in a world where your wallet can be analysed, scored and insured, the question is no longer whether you can access credit - it is what your behaviour says about you and who gets to decide.
This article first appeared in Digital Bytes (12th of May, 2026), a weekly newsletter by Jonny Fry of Team Blockchain.