Authority, autonomy, and outcomes
At FTT Fintech Festival, we had the opportunity to speak with Rohit Dhawan, Director and Group Head of AI & Advanced Analytics at Lloyds Banking Group, to discuss the practical realities of leading AI strategy, governance, and execution at enterprise scale. Rohit leads the group-wide AI strategy, sets the roadmap, and oversees execution across all Lloyds brands, spanning consumer and commercial banking, as well as investments, pensions, and insurance.
A key theme was the rapid rise of agentic AI. Rohit described it as software that is given enough authority to represent a human in two ways: the ability to handle more complex decision-making, and the permission to execute specific tasks. Agentic systems are designed to operate with greater context, navigating multi-step processes and working toward outcomes rather than simply completing pre-defined actions.
That shift raises an immediate question for any regulated organisation: how do you maintain control as systems become more autonomous? Rohit outlined Lloyds’ human-plus-AI approach, where AI is designed to work alongside people, not operate independently of them. He emphasised that, while the technology is advancing quickly, we are not yet at a point where AI can autonomously make the kinds of complex decisions organisations rely on. For now, the strongest results come from augmentation, using AI to improve productivity while ensuring human oversight, accountability, and clear operational boundaries remain in place.
Looking ahead, Rohit shared a view of the next stage of evolution: individuals supported by personal agents connected to data wallets that those agents can access with permission. In this future, agents could compare products, negotiate, and act on behalf of people in ways that are difficult today, enabling a more distributed economy. This points to both deeper collaboration and sharper competition, where products and services will need to be meaningfully differentiated and genuinely personalised to remain relevant.
Watch the full video below to hear Rohit explain “agentic AI”, the human-plus-AI approach (with people firmly responsible for oversight), and where he sees the next wave heading.
