Technology is evolving at a breakneck pace, requiring institutions to continuously adapt to remain competitive and customer-focused, says Avinash Raghavendra, Group Head – IT & Retail Operations, Axis Bank. In an interview with Christina Titus, he talks about the shift towards agentic AI, growing cyber threats, geopolitical and concentration risks arising from dependence on critical technology providers. Excerpts:
How are you navigating the rapidly evolving technology landscape?
At Axis Bank, our AI journey began well before generative AI. We have used AI and machine learning for over a decade in areas such as risk management and fraud detection. Early in that journey, we launched Adi, an AI-powered employee assistant, which has since evolved into a widely adopted platform with over 750,000 interactions and more than 35,000 monthly active users.
Today, we are building on that foundation with generative AI and agentic systems that can drive intelligent decision-making, automation and productivity at scale.
Where is AI headed next in banking?
AI is rapidly evolving from an information and recommendation engine into an autonomous execution engine. The next phase of banking will be defined by AI’s ability not only to assist decision-making but also to orchestrate actions across processes.
Our focus is on deploying AI to automate repetitive operational activities, improve service efficiency and reduce turnaround times. We are also exploring agentic AI architectures that can manage end-to-end workflows across products, operations and assurance functions, creating more intelligent, scalable and customer-centric processes.
How do you see cyber risks evolving, and what has the bank done over the past year?
Cybersecurity is undergoing a fundamental shift. Traditional perimeter-based security models are no longer sufficient in an AI-enabled world. Security must increasingly extend to every endpoint, user and machine interaction.
Over the past year, we have strengthened perimeter controls, enhanced access governance, established AI model-governance frameworks and embedded security controls within AI-assisted software-development pipelines.
Going forward, our focus will be on endpoint-level guardrails and continuous monitoring, recognising that AI tools are becoming widely accessible across the workforce.
We are also evaluating opportunities to reduce dependence on external frontier models where appropriate, while exploring greater control through internally governed AI capabilities.
What worries you most about the pace of technological change?
The pace of technological change is both the greatest opportunity and the greatest challenge. Institutions must continuously adapt to remain competitive, relevant and customer-focused.
A second concern is the growing complexity of emerging risks, particularly geopolitical and concentration risks associated with critical technology providers. As organisations increasingly integrate AI platforms into business processes, dependencies on external ecosystems become more significant.
This underscores the importance of resilience, diversification, business continuity planning and the development of sovereign capabilities that can operate independently under adverse scenarios.
How is AI strengthening cyber defence?
AI has significantly enhanced our ability to process vast amounts of information, identify patterns and respond to threats at speed and scale.
In cyber defence, this translates into faster threat detection, accelerated investigation of incidents and quicker remediation of vulnerabilities. For instance, when a zero-day vulnerability emerges, AI enables security teams to rapidly analyse intelligence, identify potential mitigations and deploy responses far more efficiently than traditional approaches. This has materially improved both our defensive capabilities and operational resilience.
How resilient are your systems to a major cyberattack?
Cyber resilience is a continuous journey. We regularly assess, test and strengthen our controls to address the evolving threat landscape. Staying alert, with no room for complacency, is key.
From a ransomware perspective, we have implemented multiple layers of protection and are further enhancing resilience through clean-recovery capabilities. We are currently validating mechanisms that can rapidly restore operations in an isolated, uncompromised environment should a major cyber event occur.
Third-party risk management is another key focus area. Dedicated teams oversee vendor governance through due diligence, ongoing assessments and site reviews, with oversight calibrated to the criticality of services and the sensitivity of data involved.
The last three questions can be added to web.
Beyond AI, what technologies should banks prepare for?
While AI will remain the dominant technology theme in the near term, quantum computing is likely to be the next major technological inflection point.
Although commercial adoption remains at an early stage, advances in quantum computing over the next several years have the potential to transform areas such as cryptography, optimisation, risk modelling and computational science. On post-quantum cryptography (PQC), financial institutions have already started thinking about implementation plans.
What steps are you taking to strengthen AI governance?
We have established a dedicated Enterprise AI organisation, led by experienced industry professionals, to drive AI adoption responsibly across the bank.
Recognising that AI is both transformative and continuously evolving, we have also created a specialised AI governance function focused on policy, risk management, compliance, model oversight and responsible AI practices.
What is your top technology priority now?
Our priority is to simplify every customer journey by making interactions seamless, predictable and friction-free, while reducing instances where customer issues are not resolved at the first point of contact.
At the same time, we are strengthening governance through greater visibility, stronger controls and robust independent oversight.