Accelerated adoption of agentic AI means business in every sector will increasingly be carried out without human interactions. That means fully-realised 24/7 commerce, as agents transact with agents with little or no need for any human intervention.
This represents a complete reinvention of the commercial paradigm, one that could deliver enormous economic upside. At the same time, however, agentic AI at scale also poses significant new risks. The solution? Trust, embedded in agentic AI systems from the outset.
Banks have a crucial role to play in building that trust. Commerce and trade are the lifeblood of the global economy, and banks sit at the nexus of commercial relationships and transactions across industries. That applies whether banks are enabling their customers to buy a property, pay for a holiday, trade on the stock market, execute cross-border payments, or supporting companies with their working capital needs.
Now, as their clients and their clients’ customers, increasingly delegate identity and execution authority to agents acting on their behalf, banks need to get on the front foot and start to rethink and reshape the central tenets of identity, access, authentication and authorisation.
Provided they do so, they will be ready to leverage the trust that they’ve built over decades, if not centuries – along with their regulated and fiduciary status – to become ecosystem orchestrators for the agentic AI economy.
Just the start of the journey
Up to now the FS industry has been making steady but measured progress in its adoption of AI. According to the FCA, more than 80% of FS organisations are already using these technologies. From unlocking speed and scalability to lowering costs and boosting productivity, the benefits to date have been significant.
But this is just the start of the journey. As the proposal for the FCA Mills Review points out: “From 2030, the AI landscape will plausibly be defined by systems that are more autonomous, adaptive, and interconnected than ever before.”
As banks and FS organisations accelerate into the agentic era, we’re already seeing significant regulatory developments. These include HM Treasury’s Financial Services AI Adoption Plan, which aims to accelerate safe AI integration while balancing consumer protection. At the same time, the Bank of England and the FCA are actively revising supervisory expectations to address operational resilience, cyber vulnerabilities, and potential multi-firm feedback loops driven by autonomous trading or credit models.
Balancing autonomy and control
Initiatives like these, that aim to boost confidence in agentic AI in FS, really matter. Recent research underlines how a lack of trust in AI is the biggest challenge facing wider adoption of bank-led AI tools. In the UK, consumers are twice as likely to believe AI will increase exposure to scams and deepfakes as they are to think it will improve financial products. The bottom line? For banks to realise agentic AI's full potential as ecosystem orchestrators, trust, risk management and transparency aren't optional extras – they're the foundations on which everything else should be built.
Agentic flows blend probabilistic AI with delegated execution capabilities which therefore are quite different to the predictable deterministic software systems in operation toady. Successful agentic systems.quire clear rules, robust safeguards and outcomes that inspire confidence. Banks must therefore be ready to manage complexity across all data, structured and unstructured. This means building contextual systems that provide the right level of control along with the appropriate degree of autonomy that makes agentic AI so powerful.
That combination of autonomy and control sits at the heart of the human + agent hybrid workforce that banks will need in order to innovate and execute workflows and processes that deliver transformational results – not just operating faster and more efficiently than before, but also exploiting entirely new possibilities that only arise when agents provide near-unlimited capacity.
Reimagining the bank’s role in an agentic world
The surge in agentic AI provides an exciting opportunity for banks and FS organisations to reimagine the role they’ll play in the future. To bring this to life, think about the process of buying a home. Today, it requires multiple complex interactions across many different providers. It’s labour intensive, costly and often frustrating for the would-be homebuyer.
In an agentic world, it all looks very different. The bank’s role switches from product manufacturer (i.e. home loans) to ecosystem orchestrator delivering embedded and intelligent outcomes. That means connecting all the parties and data involved, with trusted agents interacting and transacting with each other to help people find, finance and buy the home they want. All of this happens seamlessly with minimal input required from the customer or human servicing agents.
Trust holds the key
Consumer trust will, of course, be key for banks to be able to occupy this new ecosystem orchestrator role. Consumers will be likely to adopt AI faster for non-monetary actions, whereas the trust required to use it for financial transactions will take time to evolve and deepen.
But it’s not just a matter of consumer trust. Banks and FIs also need to trust their own technology as they embrace AI and move from experiments to enterprise-scale adoption. A key part of this will be to identify and address the risks arising from agentification. As fraudsters increasingly seek to embrace AI in their own strategies, banks will need to reimagine how they handle exceptions in this new world. That may mean enhancing their fraud detection and prevention tools, developing additional controls and harnessing AI for proactive monitoring capabilities so that ‘good’ AI can be used to tackle ‘bad’ AI.
A recent announcement from Amex, which launched industry-first agent purchase protection, is a great step in the right direction. It signals that Amex believes in their AI agents and will also build consumer confidence in the market. And that’s what will differentiate banks at the heart of ecosystems across any industry or sector.
Think strategically, act with urgency
AI will fundamentally change the nature of commerce. It will also transform how banks deliver change within their own operations, moving the focus from projects with end-dates to delivering capabilities measured by cost of ownership over time.
At Cognizant, we help FS organisations to capture first-mover advantage in agentic AI. From banks and capital markets firms to insurers and payments businesses, we’re working with them from front to back to deliver agentification at scale – refining and reinventing data architectures, workflow design, and operational governance, as well as enhancing user experiences and engineering new customer journeys. To find out how we can work with you, please get in touch.