Playing doubles with AI: How human judgement becomes the winning partner

LONDON: As businesses accelerate the adoption of artificial intelligence (AI), industry leaders have stressed the importance of human judgement, trusted data and strong governance in turning technological capabilities into meaningful business outcomes.The issue was central to Elevate Connect London 2026, the flagship event organised by Mumbai-based IT multinational ADROSONIC for senior leaders in the UK insurance market. Held at the Four Seasons Hotel London at Tower Bridge, the invitation-only gathering focused on the theme “AI Predicts. Humans Decide.”Discussions covered AI adoption, data quality, business transformation and leadership, with speakers emphasising that technology alone cannot deliver value without human expertise and organisational readiness.‘AI should complement human capabilities’One of the key sessions was a fireside conversation titled “Playing Doubles with AI”, featuring Mayank, founder and CEO of ADROSONIC, and Hasani Jess, group CTO of Aventum Group.Drawing on the example of tennis legend Leander Paes, Mayank compared effective human-AI collaboration to a doubles partnership, where complementary strengths matter more than identical capabilities.“Great doubles partnerships are not built by finding someone identical to you,” Mayank said. “They are built by combining complementary strengths and compensating for one another’s weaknesses.”He cited Paes’s career to illustrate how ambition, context and human judgement can defy expectations based solely on data. While AI can process information, conduct research and improve efficiency, people must interpret that information and decide how to act on it.Jess said effective AI use requires practice and expertise, despite the technology’s apparent accessibility.“AI is intoxicatingly familiar,” he said. “It looks easy because we can all interact with it. But just as watching a professional athlete is very different from performing at that level, working effectively with AI requires discipline, learning and experience.”The speakers also discussed the importance of trust, noting that confidence in AI systems must develop through repeated use and demonstrated reliability. With models evolving rapidly, organisations must continually reassess how best to combine human and machine capabilities.For Mayank, the real opportunity lies in using AI to achieve outcomes that would otherwise be beyond an organisation’s reach.“If AI is only helping us do the same things faster, that is one plus one equals two,” he said. “The real opportunity comes when AI enables us to do things we could not do before. That is when one plus one becomes eleven.”Business context and governance key to scaling AIAnother panel examined how businesses can expand AI adoption while protecting human expertise, governance and client trust.James Welch, head of delivery at Miller, said AI deployment must reflect the risks associated with individual business functions.While high-risk areas such as claims and payments require strong human oversight, tasks such as generating renewal packs could benefit significantly from automation. He added that protecting the company’s brand, expertise and client relationships remained a priority.Nitin Agrawal, lead business architect at ADROSONIC, said adoption, rather than technology, was often the biggest challenge.He stressed the need to redesign operating models, upskill employees and involve business users in developing AI initiatives.Daniel Del Moro, director of AI and automation innovation at UiPath, highlighted the importance of reliable data and well-defined processes.He cautioned against treating AI as a cure-all, arguing that businesses must first address process inefficiencies and data-quality problems. Combining established technologies such as robotic process automation (RPA) with newer AI tools could help unlock greater value, he said.The panel’s broader message was that AI adoption should be treated as a business transformation programme, supported by leadership, change management, clear accountability and end-to-end objectives, rather than as an isolated technology project.How much Data Quality is enough for AI?Data readiness was another focus, with leaders from AUB Group, Hampden Risk Partners, Oneglobal Broking and ADROSONIC discussing how businesses can establish reliable data foundations without delaying AI projects in pursuit of perfect information.The panel agreed that data should be trustworthy, understood and fit for the decision it is intended to support.Raj Rai said competitive advantage would depend not merely on adopting AI, but on understanding data quality, lineage, governance and how the technology is applied within an organisation.Ben Ormerod noted that the level of data quality required depends on the decision being made. Although AI can accelerate business outcomes, poor-quality inputs can also speed up poor decisions. He also highlighted AI’s potential to identify data-quality issues and trace data lineage.Marc Steel said AI should support rather than replace human expertise, particularly in relationship-driven sectors such as specialty insurance, where client relationships and market context remain important.Kamlesh Yadav cautioned that confidence expressed by AI systems should not be mistaken for accuracy. He emphasised the need for trusted information, clear ownership, human oversight and measurable business outcomes.The discussion highlighted the need to balance experimentation with safeguards, ensuring that innovation does not come at the expense of security, accountability or sound decision-making.Polar explorer draws parallels with business leadershipThe event also featured a keynote by polar explorer Ann Daniels, who drew on her expeditions across the Arctic and Antarctica to discuss leadership, resilience and decision-making under uncertainty.Daniels described operating in environments where conditions can change without warning, resources are limited and decisions can have serious consequences. In such situations, success depends on preparation, teamwork and the ability to adapt when plans no longer work.Her experiences offered parallels with the challenges facing businesses as AI capabilities, market expectations and competitive pressures evolve. Just as polar explorers must reassess conditions and adjust their routes, business leaders must make decisions without always having complete information.Daniels emphasised that technology and data can support decisions, but leaders still need to judge when and how to act on the information available. Her keynote reinforced the event’s broader focus on adaptability, collaboration and human judgement during periods of change.A Boardroom Live session also examined a proposal to expand ADROSONIC’s Digital Innovation Lab at BIT Mesra.Across the sessions, speakers returned to a common theme: AI’s value depends not only on what the technology can do, but also on the quality of the data, processes and decisions surrounding it.As organisations move beyond experimentation, effective adoption will require combining technological capability with human expertise, clear governance and trust.You use AI every day. Now get your AI Quotient. Take the AIQ test.