Every year, as banks close their planning cycles and industry observers release their forecasts, the sector performs a familiar ritual. Predictions help frame the conversation, align teams, and provide a working hypothesis about the forces that might shape the coming year. They serve a purpose.
But 2026 is unusually difficult to capture through prediction alone. The machinery of banking is shifting at a structural level. AI is advancing faster than oversight frameworks. Platforms are shaping financial decisions faster than channels can evolve. Customer expectations across retail, corporate, and Micro, Small & Medium Enterprise (MSME) banking are advancing faster than legacy product cycles. New forms of financial infrastructure, including tokenized assets and programmable liquidity, are emerging faster than most core systems can adapt.
The problem is not that 2026 is unpredictable. It is that the year is overdetermined. Multiple structural shifts are moving simultaneously, making a neatly packaged set of predictions feel too narrow for the moment.
In periods of profound transition, progress is determined far less by the ability to forecast and far more by the ability to ask the right questions, the ones that reveal tensions, clarify choices, expose assumptions, and illuminate what must be decided before the future can be shaped.
For this reason, our 2026 outlook takes a different form—one designed not to predict the future, but to sharpen the decisions that will define it.
This year we offer ten foundational questions that every banking leader, in retail, corporate, and MSMEs, will need to grapple with in 2026.Each question is built on observable signals from leading institutions, regulators, technology providers, and ecosystems.
The Ten big questions for banking in 2026
How much autonomy should banks give to AI, and how do they govern autonomous systems?
AI in banking has crossed a threshold. What began as predictive analytics and chatbot assistance has matured into agentic systems that can initiate workflows, make context-dependent decisions, and influence outcomes across credit, compliance, operations, and customer engagement. This is no longer confined to labs or pilots; it is now happening in production at scale.
In retail banking, Capital One’s Eno now handles millions of customer interactions annually, evolving beyond a basic assistant into a problem-solver that can surface relevant actions and detect unusual behavior. Across Europe, large banks such as Lloyds Banking Group are scaling AI across customer service and operations, using voice and interaction analytics to understand intent and route customers more intelligently.
Corporate and commercial banking is undergoing the same shift but with greater operational weight. Institutions in markets such as Singapore and the UAE are deploying AI to monitor covenant compliance, identify anomalies in complex entity networks, and support intraday liquidity forecasting before humans detect patterns. Treasury functions in parts of Asia are piloting AI agents that simulate scenarios and recommend how to pre-position liquidity ahead of expected market movements, pushing the boundary between decision support and autonomous action.
For MSMEs, autonomy emerges from the outside in. AI-driven accounting and invoicing platforms such as Xero, QuickBooks, Zoho, and Shopify increasingly generate the cash flow forecasts, invoice risk scores, and tax projections that banks use to underwrite credit. In many cases, the SME’s financial “brain” already sits in an AI system that the bank does not control.
Regulators are adjusting accordingly. The EU AI Act introduces obligations for autonomous and high-risk systems, including requirements around logging, oversight, and human intervention. MAS’s FEAT principles focus on fairness, ethics, accountability, and transparency in AI-enabled workflows, not just model transparency in isolation. The UK FCA emphasizes outcome integrity and end-to-end governance for AI-supported processes.
For banks, this demands a fundamental rethink of governance. Agentic governance must define what a system is allowed to do before deployment: which decisions can be fully autonomous, which require human confirmation, and which must remain human-led even if AI-enhanced. Autonomy becomes an architectural challenge, touching data, workflow design, product integrity, and pricing control.
The deeper shift is that AI is becoming a market actor in its own right. Research such as McKinsey’s Global Banking Annual Review 2025 and its work on agentic AI in banking argues that AI agents will increasingly negotiate pricing scenarios, recommend capital and liquidity positions, monitor compliance obligations continuously, and identify risk anomalies in real time.
Banks that proactively design for structured autonomy, balancing velocity with control, will gain operational leverage, regulatory confidence, and customer trust. Those that try to bolt autonomy onto fragmented systems may find that AI does not amplify their capabilities; it exposes their weaknesses.
What becomes the foundation of banking loyalty when customer inertia disappears?
For decades, customer inertia has been one of banking’s most reliable economic engines. Customers stayed because switching was complicated, corporates-maintained relationships because integration was heavy, and MSMEs tolerated inefficiencies because alternatives were scarce. In 2026, that landscape is shifting fast.
Zero-click interactions, embedded payments, and contextual financial services are dissolving the pain of switching. In markets like Malaysia and the UAE, digital-only banks such as Ryt Bank and Liv. by Emirates NBD offer anticipatory experiences where customers do not log in to bank, the system surfaces actions before they ask. E-commerce platforms, from Amazon to Mercado Libre in LATAM, now originate credit, insurance, and payments at the point of need.
Corporate loyalty is similarly vulnerable. Treasury teams increasingly demand API-first, real-time visibility across cash, FX, and liquidity, a level of responsiveness many traditional banks struggle to match. CFOs are deliberately spreading wallet share across multiple institutions to create redundancy and optionality, eroding primary-bank advantages.
MSMEs represent the least loyal segment of all. Their financial workflows are anchored in accounting tools, point-of-sale systems, and marketplace dashboards. Fintech platforms such as Qonto, Brex, and Airwallex offer integrated financial operating systems that combine payments, credit, invoicing, tax, and treasury functionality, making the bank feel more like infrastructure than a relationship hub.
With inertia collapsing, banks must rethink loyalty through the lens of the whole customer relationship. Loyalty is no longer just behavioral (frequency of use) or economic (interest rates, fees, rewards). It is also emotional (feeling supported during stress), relational (perception of fairness, transparency, and alignment), and contextual (being present where the customer’s real financial life unfolds).
In corporate banking, whole-customer loyalty emerges from consistent execution, transparent pricing, reliable liquidity, and proactive risk insights, not just relationship-manager rapport. For MSMEs, it comes from less admin, faster settlement, simpler credit access, and integrated financial tools. For retail customers, loyalty is increasingly driven by anticipatory assistance, seamless problem resolution, and empathetic support when life becomes difficult.
Banks that treat loyalty as an architectural capability, delivered through unified data, modular propositions, and context-aware experiences, will be positioned to strengthen relationships even as inertia fades. Those that continue to rely on friction as a retention strategy will find themselves competing with platforms that offer value, convenience, and intelligence with every interaction.
Who owns the customer relationship when AI becomes the first touchpoint: the bank, the platform, or the agent?
A decade ago, digital channels transformed customer engagement. In 2026, that transformation is being eclipsed by something more profound: the rise of AI intermediaries that become the first touchpoint for financial decisions.
In retail, OS-level assistants from Apple, Google, and Samsung are integrating financial reminders, budgeting nudges, and payment suggestions into everyday device experiences. Bank of America’s Erica now manages billions of customer interactions annually, while AI-powered assistants such as BBVA’s Blue and Lloyds Banking Group’s Athena platform show that customers will interact repeatedly, and often primarily, with AI before engaging a human. For MSMEs, the shift is even more pronounced. The customer relationship increasingly begins and ends inside business platforms. Accounting software recommends tax reserves, e-commerce platforms offer working capital at checkout, and invoicing systems surface cash flow risks before the bank sees a statement. Platforms such as Shopify, Stripe, and Intuit’s QuickBooks with Intuit Assist are building AI copilots that actively guide business decisions and initiate financial actions.
In corporate banking, relationship managers still matter deeply, especially for strategic and complex situations, but even here AI is reshaping the interaction surface. RM copilots, often built on platforms such as Microsoft Copilot and Salesforce Einstein, prepare meeting briefs, summarize exposures, identify anomalies, and draft communications before the human RM reaches out. AI increasingly determines when and why outreach happens, altering the rhythm and character of the relationship.
This raises a fundamental question: if AI becomes the primary interface, the channel that customers consult first and trust most, who owns the relationship?
Banks face a pivotal strategic choice. They can build their own intelligent agents that customers rely on as the authoritative source of financial insight. They can embed their capabilities into third-party agents, trading relationship ownership for reach. Or they can pursue a hybrid model that preserves brand presence while meeting customers where they increasingly live, in apps, platforms, and workflows.
This decision will shape distribution economics, trust dynamics, and brand relevance for years to come. In an AI-mediated world, relationship ownership is no longer guaranteed by account holdings. It must be earned through utility, intelligence, and availability at the exact moments customers need guidance.
Where will competitive advantage come from when foundation models commoditize?
As AI models become more powerful and more widely available, the strategic focus is shifting from intelligence to integration. Foundation models from major cloud providers now offer performance that exceeds many proprietary models, and the performance gap among leading models is narrowing, as highlighted in recent AI surveys and forecasts from firms such as McKinsey and Gartner. This makes it increasingly possible for smaller institutions and fintechs to deploy capabilities that were once accessible only to the largest banks.
When intelligence becomes abundant, what remains scarce?
The answer increasingly lies in architecture, speed, and trust.
Banks with unified data foundations, modular product manufacturing systems, and coherent pricing engines are positioned to turn general intelligence into governed, bank-grade outcomes. Those operating on fragmented cores or siloed data structures may find that more intelligence actually increases operational risk because decisions become harder to trace and inconsistencies harder to control. Industry work on modern digital cores and data platforms reinforces that integration and end-to-end visibility are now critical to unlocking AI value safely.
Competitive advantage will come from the ability to orchestrate intelligence across the enterprise: to connect customer data, product logic, compliance rules, behavioral signals, and pricing authority in a consistent fabric. This is where Zafin’s long-standing point of view, that architecture is the foundation for velocity, intelligence, and personalization, is increasingly validated by market reality.
Speed also becomes a differentiator. As foundation models equalize, the winners will be the institutions that can launch new propositions, adjust pricing, respond to regulatory changes, and integrate new partners in weeks rather than months. Interoperability, the ability to connect internal systems with external ecosystems, becomes another form of speed.
Finally, competitive advantage will hinge on trust. Institutions that can combine intelligence with integrity will define differentiation in the AI era.
What is the banking core in an ecosystem-driven world?
The traditional meaning of “the core” in banking is dissolving. Historically, the core represented a monolithic system, the operational center of accounts, payments, and product logic. Ecosystem banking changes this entirely. The embedded finance market, valued at $104.8 billion in 2024 and is projected to reach $834.1 billion by 2034 according to Global Market Insights, forces banks to rethink what they manufacture versus distribute.
Leading institutions across Asia and the Middle East now use thin cores for specific product lines or segments. Global banks experimenting with modular product and pricing systems are discovering that the core is no longer a single system but a set of institutional responsibilities: customer identity, balance sheet integrity, product manufacturing, risk and compliance, and pricing authority.
Everything else, including distribution, value-added services, and workflow orchestration, can happen in layers above or alongside the core.
Corporate banking accelerates the need for this shift. Corporate clients require coordinated orchestration of cash, credit, treasury, liquidity, and FX, functions that rarely sit cleanly inside a monolith. MSMEs, meanwhile, operate through business platforms that expect banking to integrate seamlessly into workflows.
In this context, the core becomes a question of boundaries, not systems. Banks must decide which capabilities must remain under direct control and which can be externalized or orchestrated across platforms. Institutions that cling to monolithic interpretations of the core will struggle to meet the speed, interoperability, and composability demands of an ecosystem economy.
What happens in 2026 when propositions not products become the primary lever/engine of innovation?
The industry is moving toward proposition-led competition, where customers choose outcomes, not components. This shift is most visible in corporate and MSME banking, where the lines between financial and operational workflows are blurring.
Corporates increasingly seek integrated propositions that combine liquidity, cash management, FX, treasury analytics, credit commitments, and real-time insights, not separate modules. Relationship-level pricing in Europe and Asia reflects this shift. Banks are pricing portfolios and commitments, not individual credit lines.
MSMEs experience propositions in even more blended ways. Accounting, payments, invoicing, tax, payroll, and insurance are merging into unified financial operating systems. The business owner does not want separate products; they want integrated workflows that reduce coordination costs and administrative burden.
Retail propositions are evolving similarly, especially around life events. Salary-day propositions, financial wellness bundles, and financial resilience support reflect the shift away from account-based thinking to outcome-based design.
Proposition-led innovation demands new manufacturing disciplines. Banks must be able to assemble configurations quickly, govern pricing and disclosures consistently, and adapt offerings dynamically as needs change. This is an architectural challenge, not a marketing one, and it reinforces the need for modular manufacturing and unified product governance.
What does accountability look like when banking becomes invisible?
As financial services embed themselves into non-bank environments, banking becomes invisible to customers, but accountability does not.
A corporate procurement portal that triggers an embedded credit line, a mobility app that bundles insurance and payments, and an invoicing platform that offers early wage access are all powered by banking infrastructure the customer never sees. But when something goes wrong, customer expectations remain the same. They expect the bank to be accountable.
The challenge is that accountability becomes multi-layered. Banks must govern risk and compliance across ecosystems in which multiple parties influence a single transaction. Regulators, including the OCC, MAS, and European authorities implementing DORA, increasingly emphasize end-to-end oversight across third-party dependencies, not just within the bank’s own stack.
For MSMEs, embedded finance is becoming the default. Shopify, Stripe, Square, and Intuit already provide credit, payments, insurance, and cash-flow tools from within business workflows. The business owner often cannot differentiate where the platform ends, and the bank begins.
This requires new governance constructs: real-time monitoring, chain-of-risk visibility across API layers, contractual clarity on liability, and customer-protection frameworks that extend beyond the bank’s boundaries. Invisible banking can only scale if accountability becomes hyper-visible.
Can banks digitize themselves faster than the world is digitizing assets?
Tokenization has moved from concept to capability in 2024-2025. Singapore’s Project Guardian pilots tokenized funds and deposits. Société Générale issues blockchain bonds. The UAE and Europe formalize frameworks through ADGM and the ECB. Corporate treasuries are testing programmable liquidity and real-time settlement.
The bottleneck is no longer tokenization. It is banks’ ability to modernize themselves fast enough to participate fully in a tokenized financial system.
Tokenized money demands new operational rhythms: continuous reconciliation instead of end-of-day batching, intraday liquidity optimization instead of overnight ladders, atomic settlement instead of multi-step processes, and real-time risk monitoring instead of periodic reviews.
MSMEs will experience tokenization primarily through faster cross-border payments, automated receivables, and smart invoicing. Corporates will experience it through programmable treasury operations, reduced settlement friction, and new liquidity strategies.
Banks must achieve continuous reconciliation, intraday liquidity optimization, and real-time settlement capabilities to participate in tokenized asset markets forecast to reach trillions by 2030-2035, yet most still operate on batch processing unsuited for atomic settlement. This is why modernization, data unification, and integration architectures are becoming strategic imperatives rather than technical ones.
Will operational resilience become a competitive differentiator rather than a regulatory requirement?
Resilience has historically been treated as a compliance obligation; a box to check. That era is ending. With DORA in Europe, Fed guidance in the U.S., and MAS’s Operational Resilience frameworks in Asia, resilience has become a governance priority. At the same time, customer expectations are evolving: they now interpret outages as failures of competence, not inevitabilities.
Corporate clients experience resilience as cash certainty, settlement predictability, and continuity of treasury operations. MSMEs experience it as storefront uptime, payroll reliability, and uninterrupted payments. Retail customers feel it acutely when digital channels fail during moments of financial stress.
In this context, resilience becomes a competitive differentiator. Institutions with multi-region failover, multi-cloud strategies, real-time monitoring, and embedded cybersecurity will be more attractive not only to customers but to ecosystem partners and regulators. As banking becomes more interconnected, the weakest link in the chain becomes a shared vulnerability.
Banks that articulate resilience as part of their value proposition, and not just their risk posture, will shape a new narrative around reliability as a driver of trust and choice.
Are banks transforming fast enough to lead, or only fast enough to avoid falling behind?
The final question cuts across all others. Transformation has long been framed as a program: multi-year plans, milestone charts, modernization initiatives. But 2026 requires transformation as an operating model, a continuous capability, not a project.
Agentic operations are reducing cost bases faster than anticipated. Modular architecture enables new propositions and partner integrations at unprecedented speed. Ecosystem-native competitors, many of them non-banks, are reshaping expectations across retail, corporate, and MSME segments.
Banks must now confront whether their transformation agenda is defensive or ambitious. Are they modernizing in order to maintain viability, or are they fundamentally rethinking how they create value, orchestrate ecosystems, and design the future of financial infrastructure?
Institutions that embrace clarity, speed, modularity, and disciplined execution will lead. Those that focus on incremental steps risk falling into a new gap, not between digital and analog, but between composable and constrained.
The next decade of banking will be defined by the institutions that make bold architectural and operational choices now.
Conclusion: Using questions as a strategic compass for 2026
The future of banking cannot be predicted. It can, however, be navigated with clarity, and clarity begins with sharper questions.
These ten questions illuminate where banks must redefine autonomy, loyalty, ecosystem participation, tokenization readiness, resilience, and organizational speed in 2026.
2026 is not the year to declare what comes next. It is the year to choose which questions will shape what comes next.
Banks that confront these questions with urgency and architectural discipline will define the decade ahead.
