A Decisive Moment for AI in Banking
AI in banking has moved from experimentation to a decisive phase where technology choices will shape competitiveness for years. Bain and Temenos research underlines that banks leading this transition are modernising their cores, adopting cloud-native architectures and building robust data and security foundations. These elements are critical to scaling AI safely and reliably across front, middle and back offices. At the same time, another global study highlights a widening gap between financial institutions and their supervisors in AI adoption. Financial firms are embracing advanced models far faster than regulators, creating a mismatch in capabilities and visibility over emerging risks. Together, these findings point to a sector under intense pressure: banks must pursue AI-driven innovation and banking automation aggressively, while ensuring governance, resilience and regulatory alignment keep pace with the technology’s rapid evolution.
Hyper-Personalisation and New Revenue Opportunities
Bain Temenos research identifies AI-enabled hyper-personalisation as a central growth lever, especially in retail and SME banking. By combining behavioural data with advanced models, banks can deliver more relevant, real-time product offers, improving customer engagement and loyalty. Today, banks average just 2.59 products per customer, underscoring the untapped opportunity to deepen relationships and increase share of wallet. This marks a strategic shift from simply digitising existing services to actively monetising customer experiences through AI. Banking automation, powered by cloud and governed data, makes it possible to tailor pricing, recommendations and service journeys at scale. Institutions that orchestrate these personalised interactions consistently—while protecting data and aligning with regulation—are likely to differentiate themselves in a crowded market, transforming traditional, product-centric models into dynamic, customer-centric ecosystems.
From Core Modernisation to Stablecoins and Beyond
According to Bain Temenos research, banks that succeed with AI in banking are not merely layering new tools onto legacy infrastructure. They are modernising core systems, transitioning to cloud-native platforms and establishing rigorous governance over data and security. This architectural shift enables faster experimentation, smoother integration of AI services and more resilient operations. Alongside core modernisation, Bain highlights the growing role of stablecoins in mainstream finance. While not expected to replace traditional systems soon, stablecoins are being explored for cross-border transactions, liquidity management and wholesale banking use cases. Their adoption introduces new efficiencies but also integration and regulatory complexities. Together, these trends reflect a broader transformation in financial services, where AI, digital assets and modern infrastructure interact to redefine how value is moved, stored and managed across global banking networks.
Regulatory Blind Spots and Third-Party AI Risk
While banks accelerate AI adoption, a major study involving 350 financial institutions and 130 authorities finds regulators significantly behind. Only a small share of supervisory bodies report advanced AI adoption, and just 24% collect data on industry AI use, with 43% having no plans to do so in the next two years. This lack of visibility creates an empirical blind spot that undermines effective oversight of AI in banking. The research also flags concentration risk: 69% of all respondents rely on a single leading AI provider, rising to 76% among industry users. Such dependence on a handful of powerful vendors introduces critical third-party risk, from resilience vulnerabilities to potential supply disruptions. As frontier models like Anthropic’s Mythos emerge, capable of exploiting software vulnerabilities at scale, regulators face mounting pressure to upgrade their own AI capabilities and frameworks.
The Future Outlook: Autonomous Agents and Shared Responsibility
Looking ahead, both research streams suggest AI will become deeply embedded in banking automation, from hyper-personalised customer journeys to real-time risk management and cyber defence. The Cambridge-led study argues that regulators may need to adopt agentic AI—systems capable of taking actions without constant human oversight—to effectively supervise increasingly autonomous banking models. Yet this raises complex questions about accountability when AI is built or supplied by third-party vendors. Regulators currently insist that financial institutions remain responsible for harms, including cyberattacks, regardless of where AI originates, but applying this principle becomes harder as autonomy increases. Future leaders in banking will likely be those that integrate AI and cloud technologies into every layer of their businesses, while investing equally in governance, resilience and collaborative engagement with regulators to manage systemic risk.