About this panel
Thank you to SPP Media for inviting me to join this panel in Cyprus, straight after the individual presentations, including my own keynote, "Who Wins in 2040?" It was a privilege to share the stage with H.E. Mohamed Rashid bin Taliah, who shared the UAE's experience of using agentic AI to transform the public sector, Deputy Minister Nicodemos Damianou, who connected the discussion to Cyprus's national AI, innovation and digital transformation priorities, and George Tziortzis, Head of Consulting Services at EY Cyprus, who brought the view from government and business transformation work across Europe. Thank you too to Michalis Sophocleous for moderating.
Serving people, not deciding for them
Michalis opened with the elephant in the room: if agentic AI acts on behalf of the state, does the state end up deciding for people rather than serving them? His Excellency was clear that in the UAE it is a hard rule that the systems do not make decisions on behalf of people. Every action that needs a citizen's consent is approved by that citizen first. What the agents do is take on the jargon and the paperwork, so the state gives people time back instead of taking it from them. And the UAE is not transforming everything at once. It moves one service at a time, lets people decide whether it makes sense, and adjusts as it goes.Bureaucracy is not the same as digitisation
Deputy Minister Damianou was candid about the challenge for Cyprus. Too often we confuse bureaucracy with digitisation, and putting a slow process online does not make it fast. The public service still rewards people through a strictly hierarchical model when it needs to reward creative thinking, critical thinking, decision making and adaptability. Cutting bureaucracy needs a whole of government and whole of state effort. Cyprus will not compete on scale, but it can do a great deal by helping people, SMEs and businesses use the AI tools that already exist to raise productivity.The panel also came back to how governments buy technology. A procurement cycle of eight to twelve months just to start a project defeats the whole idea of moving at the speed of AI, and that needs to change. Traditional governments were built around their regulatory and legislative environments first, not around the citizen, which is why the mindset shift matters so much.
Where the human sits in an AI native organisation
I shared how the 311 Institute works as a fully AI native company. Most organisations bolt AI onto a digital business and end up with a Frankenstein. This year I spent six months tearing down the company I had spent ten years building and rebuilt it from scratch. During the event I took the Deputy Minister's digital zero bureaucracy ideas, turned them into an internal policy, and had our systems start identifying the bureaucratic steps in our own processes, including the websites people were using in the room. Every new policy then goes through our governance, risk and compliance layer, with internal and external auditing, because adopting someone else's strategy only works if you check it is right for you and understand the new risks.On where the line sits between people and agents, our stance is simple: we do not trust AI. We check every input and every output, and trace everything back to the data. Agents take low risk, tactical decisions, where low risk means a small blast radius if something goes wrong. People make the strategic decisions: the vision, the mission, the strategy, and what we build and for whom. And we keep a complete audit trail of every decision an agent makes, so we can review them and step in where we disagree.
From pilots to end to end
George, fresh from events in Stockholm, Athens and Malta, said the message everywhere is the same: the mindset has to change fully. Plugging AI into old processes only creates more patching. The next step is to move from pilots to leapfrogging, building services end to end with AI. The public sector's usual problem is execution and the speed of deciding, yet there are already strong examples to learn from, including digital assistants, early warning systems that predict problems well in advance, the European principle of single submission, and using AI to work across fragmented government data instead of trying to pull it all into one big data warehouse.A hopeful message for young people
Michalis closed by asking each of us for a hopeful message for the many young people in the room. Mine was that education has taught us to play defence: get a skill, get a job, and stay there for fifty years. Today anyone can use AI that gets more capable every month to start a business or multiply their own work many times over. So go home tonight, open any AI tool, and ask it in plain language to help you build a simple app, for your homework or your football scores. The people who are creative, ambitious and know how to use the technology are the ones who will benefit, while those who do not risk falling behind in a K-shaped economy.His Excellency added that nobody is born knowing algebra: start with a first experiment, such as a simple bouncing ball game, and you will learn fast. That applies to people and to governments. The Deputy Minister pointed to Cyprus's new strategy and its relentless focus on adopting and applying AI and preparing people for the intelligence age. And the panel ended on a hopeful note: this is a stressful period of transition, there will be failures along the way, but people can bounce back faster than ever before.
