Anu: I am Head of Programme Delivery at NHS Blood and Transplant. I have worked in the technology and digital space for over 22 years. Earlier in my career I worked at the Indian Space Research Organisation and at the Indian Institute of Science, so I bring a strong science and engineering foundation. About three years ago a friend referred me to this role at NHS Blood and Transplant. It felt right because of the clear purpose of saving and improving lives. I have lived those values since joining and I love what I do.

Interviewer: Based on your experience leading large scale digital transformation, what are the key factors for a successful AI initiative?

Anu: People often confuse technology readiness with organisational readiness. Technology readiness is usually easier to achieve. Organisational readiness is harder and where many efforts stall. To succeed, you need a clear view of the value AI will bring, the cost, and whether the organization is prepared to accept and move in that direction. If you do not have that alignment, the initiative will struggle.

Interviewer: Your session explored what comes after large language models. What emerging systems or technologies do you see as the next frontier?

Anu: Today many people use AI like an information or knowledge base. You ask a question and it returns information. I see AI moving from something we talk to toward something we collaborate with. Agentic AI is gathering momentum and I expect multi agent autonomous decision making to become more common. I also expect more small language models, trained on specific context and proprietary data, which are especially appropriate in highly regulated environments. People will become more AI literate and more accepting of AI at work. One analogy I like is from the industrial revolution. Many jobs changed from people needing to shower after work to needing to shower before work. With the AI revolution we will work differently from today. We may not yet know exactly how, but we should be open to that change.

Interviewer: How should large organisations think about this shift?

Anu: Go back to foundations. Look at your core processes, your data, your integrations and your underlying intelligence. Make sure those are robust. If they are not and you just add agents and models on top, you are only automating confusion. That will not work for anyone. Strong foundations matter.

Interviewer: The NHS operates within strict regulatory frameworks. How do you implement AI responsibly while ensuring patient safety and measurable outcomes?

Anu: The first question is whether AI is needed at all in the specific context. Sometimes what you actually need is targeted automation, not AI. Because we work in a highly regulated environment with patient safety at stake, we are very cautious. Some processes are still on paper for good reasons. If AI is appropriate, the next question is the level of autonomy. Do we need a human in the loop or on the loop? Should we limit AI to operational tasks like compiling reports and following up actions, and stop short of anything that could affect clinical decisions or patient outcomes? We want to keep human decision making in place. Responsible AI implementation is essential.

Interviewer: How do you stay human centered in healthcare, where patient outcomes and ethics are paramount?

Anu: Take the example of a donor journey. Using AI to send reminders or offer a pass on a phone so a donor can check in more easily is positive. But using AI to manipulate or unduly influence decisions crosses an ethical line. We should not do that. This is a pivotal moment. As leaders we must ensure ethical, responsible AI and we must speak up when we see risks. It is our responsibility to get this right.

Interviewer: What should leaders focus on if they want to pursue AI ambition and keep innovating?

Anu: Leaders do not need to be AI engineers. They do not have to build models. But they do need AI literacy. They should understand responsible implementation, guardrails, governance, compliance and policy. They also need a strong grasp of organisational processes and to keep foundations strong. Maintain critical thinking and act in the organisation’s best interests. Aim for responsible AI that is sustainable and acceptable. Most importantly, take your people with you. The secret to success is adoption within the organisation.

Interviewer: You are a STEM ambassador and an advocate for inclusive leadership. How do diversity and inclusion shape AI outcomes?

Anu: It is crucial to capture diversity in the data we use. If biases exist in foundation data, those biases will amplify as we integrate systems and build AI layers on top. Inclusive leadership brings diverse voices and diverse data into the process. When AI is built on that, the outcomes reflect more perspectives. Diversity and inclusion are as important as the technology and as organisational adoption and readiness.

Interviewer: Looking ahead, what trends do you expect to see in healthcare over the next 12 months?

Anu: Small language models trained on proprietary data in regulated environments will accelerate. Multi agent environments will advance. We will also see progress in robotics for assisted living. The key questions are how ethical and how responsible these deployments will be. We need to ensure they are both. As leaders we should raise our voices when they are not and take responsibility for making sure they are done right.

Everyone is moving at 100 miles an hour with AI. Some are adopting it because it’s a shiny new thing, others because it’s the right thing to do, and many because it will make life easier across the organisation.

That’s why I come to The AI Summit London: to hear from those who’ve actually implemented AI—how they did it, the lessons learned, the pitfalls, and what they’d do differently if they started again.

The learning opportunity these gatherings provide is invaluable. Today has been fabulous: I’ve met some brilliant people, had genuinely enlightening conversations , and I’m looking forward to the rest of the day. If you’re serious about AI, you should be here.

Closing

From foundational readiness to responsible adoption, Anupama Hatti emphasises that the next phase of AI in healthcare is as much about people, process and ethics as it is about technology. 

Expect growth in small domain specific models, multi agent systems and assisted living robotics, guided by strong governance and a commitment to patient safety. The opportunity is to collaborate with AI while keeping human judgment at the core.