Dynamic Democracy in the age of artificial intelligence

In last week's article, we discussed planning and the price mechanism in the age of artificial intelligence. Artificial intelligence can ensure the effective and rational functioning of complementary planning and market mechanisms. However, more powerful computers and more advanced algorithms alone will not be enough for this. 

Planning means making choices. Making choices means political economy, it means politics. For this, we need to be in search of a strong democracy.  

Following last week's article, our esteemed professor Prof. Dr. Bilsay Kuruç sent a note. In summary, he says: 

“From today on, there are two 'must-haves' for us to be able to talk about a Plan. First, politics. Today, this does not exist at the political level in our country!

Second, the world of economists… Our colleagues have been living in an 'interest rate-exchange rate prison' for a long time. They have gotten used to it. I cannot see anyone who has left that place to start working on a 'Plan', who is eager to start, or who puts forward views at the negotiating table 'starting from the ground of theory'.”

Keeping this warning from our professor Bilsay Kuruç in mind, let us continue to discuss the political dimension of planning and democratic choice mechanisms in the age of artificial intelligence.

Who will make the choices of the plan? Who will oversee those who make the decisions? Is what we call democracy just going to the ballot box every four or five years to elect administrators?

Is a “Dynamic Democracy” possible in the age of artificial intelligence where citizens can continuously participate in decision-making and oversight processes?

Artificial intelligence can help us calculate the potential outcomes of options and our choices. But it should not make the choice. What kind of growth, what kind of distribution? More consumption or more leisure time? Defense or healthcare? The welfare of today's generations or the resources we will leave to future generations?

IS DEMOCRACY JUST THE BALLOT BOX?

Here we return to the problem of democracy. Free elections are an indispensable condition of democracy, but democracy is not just elections. Voting every four or five years and delegating all the country's choices to elected officials until the next election cannot be a necessary but solely sufficient definition of democracy.

The ballot box, parliament, independent judiciary, accountable executive, a free press not held captive by capital monopolies, unions, professional organizations, universities, non-governmental organizations, cooperatives... All of these are the mutually checking parts of the democratic system.

Power does not gain legitimacy only through elections. The authority exercised by the government must also be constitutional authority and must be subject to constant oversight.

In the third article of the Utopia and Essays series, we recalled the question that the whole world is debating:

 Who will control artificial intelligence? Who will control those who control it?

The same question applies to political power as well. Especially in today's world where authoritarian tendencies are increasing, what mechanisms will control political power?

Artificial intelligence has begun to increase the state's capacity to collect information, analyze, make decisions, and implement them to an extent never seen in history, and it appears that this will increase even further. Along with this, the citizen's capacity to oversee the state will also increase. What matters is whether the citizen can use this new power of oversight.

A study I encountered while preparing this article addressed the problem we are discussing from another angle.

Prof. Steve Jacob from Laval University in Canada, in his article titled “Artificial Intelligence and the Future of Evaluation: From Augmented to Automated Evaluation” published in 2025, examines how artificial intelligence could change the evaluation of public policies. (1)

We can roughly think of today's classic public policy cycle as follows:

A problem arises › The state defines the problem › Develops policy › Makes a decision › Implements it › Measures the results › Evaluates › Changes the policy if necessary.

But this process sometimes takes years. By the time it turns out that the policy is wrong, resources may have been spent, and the economic and social consequences may have already emerged. Artificial intelligence can create a significant change here.

From production to consumption, from education to health, from traffic to energy, from prices to employment, vast amounts of data are constantly being produced. By evaluating this data, artificial intelligence can enable us to analyze the results of an implemented policy in a much shorter time.

Policy is implemented › Data arrives › Results are evaluated › Policy is corrected › New data arrives › It is re-evaluated.

Thus, policy-making can transform from a linear process into a continuously operating feedback system.

This approach is called the “Dynamic Public Policy Cycle” in academic literature. David Valle-Cruz and his colleagues examined the potential of artificial intelligence to make the public policy cycle more dynamic in 2020. Steve Jacob, in his 2025 study, addresses this approach with the concept of the “Dynamic Public Policy Cycle”. (2)

The OECD report titled “Governing with Artificial Intelligence” published in 2025 also uses the same concept. The OECD discusses that with evaluations becoming faster and lower-cost thanks to artificial intelligence, it is possible to transition from the classic system where results reach decision-makers too late, to a system where policies can be shaped, corrected, and redesigned in almost real-time. (3)

IF POLICY IS BECOMING DYNAMIC, WHY SHOULDN'T DEMOCRACY BECOME DYNAMIC?

The state constantly evaluating its own policies can make administration more effective. But the effectiveness of administration does not mean the strengthening of democracy. If the state can constantly monitor the results of the policies it implements, evaluate them with artificial intelligence, and change them, where will the citizen be in this process?

Artificial intelligence predicting what people think from social media messages, consumption behaviors, internet searches, and hundreds of other data sources is not democracy. At most, this creates a government that has more information about society and opens the doors wide to the authoritarianism and digital fascism we fear.

Democracy, on the other hand, is the citizen being able to express their own will. Expressing it is not enough. They must be able to discuss, access information, see alternatives, participate in the decision-making process, and oversee the results of the decisions made.

In the same OECD report, artificial intelligence and citizen participation are also addressed separately. It is discussed that digital technologies can create new channels of participation between citizens and public institutions; and that artificial intelligence can classify and summarize large amounts of opinions and increase the scale of participation in the preparation of public policies. (4)

The study titled “Artificial Intelligence and the Future of Citizen Participation” published by the OECD in 2026 takes this discussion even further. Examining 50 applications from 22 countries, the study shows that artificial intelligence can enable citizen participation and deliberation on a scale that was previously technically impossible. (5)

But there is a very important warning here: Artificial intelligence should not replace human participation. It should strengthen human participation.

The “Dynamic Democracy” discussion starts here. 

DYNAMIC DEMOCRACY

The words “Dynamic Democracy” are of course not being used here for the first time. There are studies in the political science literature that use this name with different meanings. (6)

While discussing the subject with a friend, we asked this question. 

If the state's decision-making, implementation, and evaluation system is constantly working in the age of artificial intelligence, why should the citizen's participation in democracy work only once every four or five years? Does Dynamic Democracy emerge from here? 

I am not talking about abolishing representative democracy. I am also not talking about an electronic referendum system where millions of people vote on every issue from their mobile phones every morning.

A system where citizens participate in voting on every issue can easily turn into a system of manipulation. It is also not possible for millions of people to specialize in every public policy. Furthermore, the fundamental rights of individuals, classes, and social groups cannot be made a subject of voting. What we are looking for is not to replace representative democracy with another system, but to strengthen it with continuously operating participation and oversight mechanisms.

Local and national councils, sectoral boards, citizens' assemblies, unions, professional organizations, cooperatives, universities, non-governmental organizations, and new democratic institutions that we have not even named yet. Artificial intelligence can reduce a very important technical obstacle in front of these institutions.

Evaluating the opinions of millions of people was almost impossible until today. With artificial intelligence, opinions can be classified, common demands can be determined, the results of different proposals can be calculated, and alternative costs can be explained to the citizen in an understandable way.

For citizen participation to turn into a real oversight mechanism, decision-makers must have the obligation to take these opinions into account, explain the justifications for their decisions, and open the results to public scrutiny.

Thus, democracy can cease to consist solely of the cycle of election › government › wait four-five years › re-election.

Instead, a continuously operating democratic feedback system can be established in the form of: choice › decision › implementation › data › evaluation › citizen oversight › correction › re-decision.

This is what we mean by “Dynamic Democracy” in this essay. Naturally, the separation of powers, parliamentary oversight, independent judicial oversight, and a free press that should exist in democracies must be strongly protected.

THE SAME TECHNOLOGY CAN BUILD TWO DIFFERENT WORLDS

But right here, the problem that has been following us since the beginning of the series reappears.

Technology does not tell us which political order to establish. The same artificial intelligence infrastructure can be a tool for Digital Fascism on one side and Dynamic Democracy on the other.

The technology is the same. The political choice is different!

For this reason, the fundamental question of democracy in the age of artificial intelligence will not only be “Who will govern?”

Who will oversee those who govern? Who will oversee the algorithm? Who will decide who owns the data the algorithm uses? And perhaps most importantly:

How much information will the citizen have about the state and technology companies that know more and more about them, and how will they oversee them? 

Let us assume that we have established the most advanced Dynamic Democracy in the world in a country, that we share the wealth created by artificial intelligence fairly, and that we have created an order where the citizen truly participates in economic and political decision-making processes.

What if this wealth is based on the mines, energy, data, and poverty of other countries? Can one country's utopia be built on another country's exploitation? 

Two countries stand out in the artificial intelligence race today: the USA and China. So what will happen to developing countries?

We are coming back to the point we made in the second essay:

If utopia is not universal, it is not utopia.

Then the next question is:

What form will imperialism take in the age of artificial intelligence?

In the next essay...

Notes

(1) Steve Jacob, “Artificial Intelligence and the Future of Evaluation: From Augmented to Automated Evaluation”, Digital Government: Research and Practice, Vol. 6, Issue 1, 2025, Article 10.

(2) David Valle-Cruz, J. Ignacio Criado, Rodrigo Sandoval-Almazán and Edgar A. Ruvalcaba-Gomez, “Assessing the Public Policy-Cycle Framework in the Age of Artificial Intelligence: From Agenda-Setting to Policy Evaluation”, Government Information Quarterly, Vol. 37, Issue 4, 2020, 101509.

(3) OECD, Governing with Artificial Intelligence: The State of Play and Way Forward in Core Government Functions, OECD Publishing, Paris, 2025. Especially the “AI in Policy Evaluation” section.

(4) OECD, Governing with Artificial Intelligence, 2025, “AI in Civic Participation and Open Government” section.

(5) OECD, Artificial Intelligence and the Future of Citizen Participation: Typology of Applications, Opportunities and Challenges for Democratic Innovation, OECD Public Governance Reviews, OECD Publishing, Paris, 2026.

(6) Devin Caughey and Christopher Warshaw, Dynamic Democracy: Public Opinion, Elections, and Policymaking in the American States, University of Chicago Press, 2022. The meaning of the “Dynamic Democracy” concept here is not the same as the AI-supported continuous participation and oversight model proposed in this essay.