As AI reshapes every part of the business, the lines between data and AI leadership are shifting. This panel explores how the Chief AI Officer and Chief Data & Analytics Officer work together (or diverge) to lead strategy, scale, and governance in the AI-first enterprise. 
 

• Redefining executive roles: Where the CAIO leads vs where the CDAO continues to own — and why both roles are becoming critical in AI-centric organisations 

• From enablement to ownership: How AI leaders are moving from supporting functions to owning products, P&L, and core transformation agendas 

• Merging data and AI strategy: Why decoupling AI from data doesn’t work — and how smart organisations are building integrated delivery across both 

• C-suite alignment on AI risk: Ensuring the CEO, CTO, CFO and board all understand what responsible AI looks like — and who’s accountable when things go wrong 

• Talent, culture and credibility: What CAIOs and CDAOs must do to build influence, lead change, and create a future-ready organisation 

TBC: Miles Hillier, Chief Product & Technology Officer, AI – NatWest Group 

With new laws like the EU AI Act on the horizon, organisations face growing pressure to enforce strong AI governance. This discussion brings industry leaders together to explore how to build ethical, compliant AI frameworks that scale without stifling innovation. 

• Preparing for regulation: How enterprises are gearing up for incoming AI laws (e.g. the EU AI Act), establishing internal policies and controls to meet stricter requirements. 

• Foundations of ethical AI: Defining the cornerstones of AI governance in 2026 – from transparency and fairness to accountability – and aligning AI projects with these principles. 

• Scaling compliance wisely: Strategies to expand compliance and risk management processes (model audits, documentation, oversight) in a way that doesn’t slow down AI delivery. 

• Cross-border governance: Managing AI governance in multinational organisations, ensuring consistent standards while respecting different regional regulations and cultures. 

• Accountability at the top: Embedding AI risk management into corporate governance structures so boards and executives have clear oversight of AI ethics and performance. 

Many organisations have piloted generative AI – now the challenge is turning those experiments into real, scalable value. This session shows how to embed GenAI into products, operations and customer experiences at scale, using the right tools and practices to bridge the gap from hype to impact. 

• Use cases to enterprise solutions: Examples of how companies are moving from isolated GenAI use cases to integrated solutions in customer service, operations, and product innovation. 

• Building the AI platform: Setting up enterprise-grade infrastructure and MLOps pipelines to support GenAI deployments – ensuring reliability, integration, and efficiency as usage grows. 

• Measuring what matters: Establishing clear benchmarks and KPIs to track GenAI’s performance and business impact, helping separate genuine ROI from the hype. 

• Overcoming scaling hurdles: Tackling practical challenges in scaling GenAI, from handling massive data and retraining models to controlling costs and maintaining model performance. 

• Safe and controlled deployment: Implementing guardrails for GenAI (content filters, human oversight, robust testing) so that scaled deployments remain compliant, secure, and trustworthy. 

This session examines what next-generation automation looks like when human expertise, AI and robotics combine, and how organisations can prepare for these changes. 

• Beyond traditional RPA: How advanced AI (like large language models) is automating tasks that were once beyond the reach of basic RPA, opening up new efficiency opportunities. 

• Workflow transformation: The impact of AI-driven automation on core functions – from finance and HR to supply chains – and where businesses are seeing the biggest productivity gains. 

• Readiness for autonomy: Gauging organisational readiness to evolve from scripted RPA processes to more autonomous AI systems, including the skills and mindset shifts required. 

• Architecture and ROI: Designing the right architecture to blend AI with RPA tools, and learning from real-world examples of ROI achieved through intelligent automation initiatives. 

• Human-in-the-loop: Ensuring human expertise and oversight remain integral in automated workflows, so employees can manage exceptions, maintain quality, and trust the AI-augmented processes. 

In this session, experts discuss how enterprises are leveraging multimodal AI (combining text, images, speech, and more) to unlock new capabilities and business value. 

• Leveraging all data types: How organisations are harnessing multimodal models that fuse text, vision, speech and structured data to solve complex problems and deliver richer insights. 

• Foundational vs bespoke models: Deciding between giant general-purpose models and specialised domain-specific models – understanding trade-offs in performance, cost and control. 

• New use cases emerging: Examples of multimodal AI in action – from intelligent assistants that see and talk, to image-enhanced analytics and robotic systems with combined senses. 

• Deployment challenges: Practical challenges in implementing multimodal AI at scale, including integrating diverse data sources, high computational demands, and ensuring user adoption. 

• Keeping it responsible: Techniques to maintain transparency and manage bias in multimodal AI systems, ensuring these complex models meet enterprise standards for ethics and compliance. 

TBC: Sumit Goyal, Chief Operating Officer and Gen AI Program Finance Lead - Cloud & Innovation Office – Deutsche Bank 

This interactive group discussion explores how creating artificial data can accelerate AI development while preserving privacy – and what it takes to extract genuine business value from these techniques. 

• When to go synthetic: Identifying scenarios where synthetic data adds value, such as augmenting limited datasets or enabling data sharing in regulated industries without breaching privacy. 

• Speed and privacy by design: How high-quality synthetic data can speed up model training and testing while ensuring compliance with data protection laws and customer privacy expectations. 

• Mitigating risks: Understanding the risks of synthetic data – from potential biases to lack of real-world fidelity – and how to govern and validate synthetic datasets to ensure trustworthiness. 

• Integration into pipelines: Best practices for incorporating synthetic data generation into the AI development pipeline, including tools for creating, validating and updating synthetic data alongside real data. 

• Balancing with real data: Strategies for combining synthetic and real data in production models so that AI systems remain accurate and relevant to actual business conditions. 

You can’t deliver on AI without the right people – and those people are getting harder to find and keep. This session examines how to build an AI talent strategy that can scale, covering everything from smart hiring and upskilling programmes to retaining critical experts in a competitive market. 

• Recruiting AI talent with the practical skills to deliver real-world results – looking beyond buzzworthy CVs to find candidates who can bridge technology and business needs. 

• Developing internal training programmes to cultivate an AI-literate organisation, so employees in all departments can understand and leverage AI (not just the data science team). 

• Keeping top AI professionals engaged by providing clear career paths, technical leadership opportunities, and a culture that values innovation and autonomy. 

• Making strategic choices about growing in-house AI expertise versus outsourcing or using automated tools – balancing cost, capability and long-term agility. 

• Organising AI teams effectively, from deciding whether to centralise a data science function or embed experts in business units, to aligning talent structure with the company’s AI goals. 

AI must be ethical and trustworthy across all sectors – not just in highly regulated fields. In this talk, we highlight practical steps and best practices to implement responsible AI at scale, moving beyond theory into day-to-day AI governance that applies in any industry. 

• Enterprise AI ethics: Implementing AI ethics frameworks that address fairness, transparency and accountability, and adapting these principles to fit the specific risks and regulations of different industries. 

• Audit and oversight: Techniques for auditing and monitoring AI models in production to catch bias, errors or drift early – ensuring systems remain fair, accurate and compliant in real-world use. 

• Cross-functional governance: Setting up governance teams that include not only AI experts but also compliance, legal and domain specialists, so ethical oversight is built into AI projects from the start. 

• Industry collaboration: How companies can collaborate across industries to develop common standards and share knowledge for responsible AI (through consortia, open frameworks, etc.), improving AI safety collectively. 
• Built-in responsibility: Infusing responsible AI practices into the AI development lifecycle – from careful, diverse data sourcing to transparent model design – so that ethics and safety are “baked in” rather than bolted on. 

Deploying AI in a few pilot projects is one thing – operationalising it company-wide is a far bigger challenge. This panel discusses how to turn isolated successes into a truly AI-enabled enterprise, tackling the technical and organisational hurdles along the way. 

• From pilot to platform: Moving from successful AI proofs-of-concept to enterprise-wide platforms – how to pick the right pilots to scale and avoid common pitfalls that derail broader adoption. 

• Integration at scale: Making AI work in the real world of legacy IT systems and edge devices – ensuring new AI solutions can plug into existing infrastructure without compromising performance or security. 

• Workforce enablement: Bringing employees along on the journey by providing training and change management, so staff across departments are prepared to use AI tools and adapt their workflows. 

• Operating model for AI: Establishing an operating model that supports AI at scale – including ongoing model maintenance, support teams, and governance – to keep AI systems running reliably over time. 

• Measuring impact: Tracking and communicating the business impact of AI at scale (ROI, efficiency gains, risk reduction) to demonstrate value, build trust, and guide future AI investments. 

As AI moves from pilot projects to production systems, enterprises need a reliable “AI factory” to manage models at scale. This presentation covers how to set up the right architecture, tools and processes to deploy AI securely and efficiently across the organisation. 

• Enterprise architecture for AI: Designing AI infrastructure that can support training, deploying and monitoring many models at once – across on-premises data centres and cloud – without sacrificing performance. 

• MLOps best practices: Implementing MLOps pipelines (CI/CD for machine learning) to streamline the model lifecycle, from data preparation and model development to deployment and updates in production. 

• Robust monitoring & governance: Putting in place automated monitoring to detect model drift, bias or performance issues, coupled with governance controls to ensure compliance with regulations and ethical standards. 

• Security by design: Integrating security into every step of the AI pipeline – protecting sensitive data, controlling access to models, and guarding against threats or tampering that could compromise AI systems. 

• Balancing standardisation and agility: Providing central AI platforms and tools that increase efficiency and reuse, while giving teams enough flexibility to innovate and tailor solutions to their business needs. 

For AI to truly transform the enterprise, it must move beyond siloed projects and become part of the organisation’s core strategy and culture. In this closing session, we explore how leading companies are weaving AI into every department and function, ensuring AI underpins decision-making and innovation enterprise wide. 

• Company-wide AI vision: Case studies of firms that evolved from a few AI experiments to a holistic AI strategy led from the top – illustrating what it takes to get company-wide alignment on AI. 

• Culture and literacy: Methods for embedding AI into the company culture – from AI education initiatives to champion programmes – so that employees at all levels become confident using AI in their work. 

• Strategic alignment: Ensuring each AI initiative ties into business goals and strategy, turning AI from a technical endeavour into a strategic tool for revenue growth, customer satisfaction and efficiency. 

• Central guidance, local execution: Balancing a centralised AI governance and support function with empowering individual business units to innovate – maintaining standards while encouraging grass-roots AI solutions. 

• Democratising AI access: Scaling the infrastructure and support so that AI capabilities (data, models, tools) are available on-demand to all departments in a secure, governed manner – making AI as ubiquitous as IT in the organisation.