Leading hybrid human-AI teams: Redefining roles, trust and accountability in the UAE

When I walk into a modern financial services office in the UAE, I no longer see just people. I see relationship managers, analysts and team leaders. On their screens, I also see chatbots answering customers, engines scoring risk and algorithms flagging suspicious activity. This is a hybrid team – part human, part AI.

By the end of this decade, almost every team in global organisations will look like this in some form. The important question is not whether we will lead hybrid teams, but how. Many organisations start with technology. Companies are investing in new systems, running training and hoping productivity improves. But if we don’t also redesign roles, expectations and ways of working, we create confusion. 

Who is responsible?

I have heard it in many conversations with colleagues: ‘Is this tool here to replace me?’, ‘Am I responsible if the system makes a mistake?’, ‘Can I challenge the recommendation?’ In culturally diverse teams, these questions can be even harder to voice openly.

These concerns are not theoretical. Let us take an example from financial services. Imagine a lending officer in the UAE who approves a borderline case, guided by an AI model. The customer later defaults. Or a compliance analyst who dismisses an automated alert that turns out to be important. Who is accountable in these scenarios – the individual, the team leader, the vendor or the regulator? If this is unclear, trust quickly erodes. People either hide behind the system (‘It wasn’t my fault’) or distance themselves from it (‘I never really trusted it’). Neither response helps the organisation learn.

Leading hybrid teams means shifting from ‘I manage people using tools’ to ‘I orchestrate a system where some team members are algorithms’. That sounds abstract, but it leads to very practical questions:

• Which tasks belong to humans? 
• Which tasks are shared with AI, and which can be largely automated? 
• When human judgement and AI outputs are in conflict, which must I believe? 
• How do we encourage people to speak up when they feel the system is wrong? 
• How do we make sure that we are not embedding bias or unfairness into our automated processes?