It seems to be popular to speak about Artificial Intelligence (AI) and its consequences. But what is AI? There are many definitions of Artificial Intelligence. Let‘s just use the first one popping up in the Google search function: It defines it as “the application of computer systems able to perform tasks or produce output normally requiring human intelligence, especially by applying machine learning techniques to large collections of data.“ Another definition I found in Mustafa Suleyman’s book “The Coming Wave – Technology, Power, and the 21st Century’s Greatest Dilemma”. Suleyman defines AI simply as “the science of teaching machines to learn humanlike capabilities.”
AI – a civilizational technology
Already the definition of AI and especially the title of Suleyman’s book show us into which kind of dilemma we move, when thinking and writing about AI. One of the dilemmas is the fact, that we do not know today if the technological development will in future also lead us to what is called Artificial General Intelligence (AGI), which Suleyman defines as the “point at which AI can perform all human cognitive skills better than the smartest humans.” In this connection one speaks also of a super- intelligence AI. We are not there yet, and we do not know if we will ever see it appearing. But it seems not impossible that humankind reaches that point.
Anyway, looking at the present stage of achievement and thinking about possible developments it is clear our society is confronted by enormous technological challenges. One of the top AI experts Dr. Fei-Fei-Li, a Chinese immigrant into the U.S.A., speaks about “civilization technology”, because everybody will be effected by AI. It will change in dramatic ways, how our society is functioning. Therefore, it is absolutely necessary to study the effects of AI on all issues of our societies and not least to look into the influence of AI on our democratic systems. We have to study if and how AI is endangering democratic values and principles. And vice versa, we have to ask if AI can be used to strengthen our democracies.
Mark Coeckelbergh, Professor of Philosophy of Media and Technology at the University of Vienna, deals with this question in his recent book under the title “Why AI Undermines Democracy and What To Do About It”. As mentioned above, Dr. Fei-Fei-Li speaks about a “civilization technology” and the Israeli historian Yuval Noah Harari speaks about the end of human or better “human dominated history”. With AI we got – and with superintelligent AI it would be even more the case – another being who is defining the course of history. It will be a hybrid world, where the human and the virtual world will mix. What does that mean for our human community?
Influencing AI’s development
Some even fear that the development in the direction of Artificial General Intelligence could lead to the destruction of humanity. Therefore, they appeal to the governments around the world to ban this development. In a recent edition of the American magazine “Time” you could find such an appeal under the title “THE BAN WE NEED”: “No one knows how to control AIs that are vastly more competent than any human, yet we are getting closer and closer to developing them, with many experts expecting superintelligence in the next five years, at the current pace. This is why leading AI scientists warn that developing superintelligence-AI that outperforms human across all cognitive tasks – could result in humanity’s extinction.”
But already now we are confronted with several dangers for our democratic systems of governing. One of the important issues we have to deal with is AI’s enhancing of already existing social deficiencies. Many authors studying and analyzing our societies and particularly the influence of social media, notice a growing isolation, especially amidst the younger generation. This isolation due to the concentration on media related communication is undermining and at the end destroying the direct communication between humans, which is necessary for a viable democracy. Democracy is depending on communication and exchange of ideas and opinions in order to find compromises. If such a minimum of direct exchanges – also between people of different ideas – is substituted by “communication” inside bubbles it undermines democracies. Mark Coeckelbergh says it bluntly: “Isolation kills”, it kills direct communication and with that it kills democratic societies.
Artificial Intelligence goes a step further. It reduces even more the necessity of communication. You may observe it in the education system, where AI is delivering elaborated theses and essays. And the danger is, that with these texts ideologies are delivered as a by-product. And these by-products can be even more dangerous than some mistakes or fake news as part of the content offered. The ideological by-product can directly offer values and principles which are putting democratic values into doubt. It can also have a negative influence by omitting important values and principles on which democracy is built.
Henry Kissinger, Craig Mundie and Eric Schmidt speak in their book “Genesis” of an “unique capacity in human psychology en masse”. They fear that “an AI could also hijack a rival nations media, producing a deluge of synthetic disinformation”. And this could also be “so alarming as to inspire mass opposition against further progress on that country’s AI capacities”: a war between AIs. Decisive for democratic concerns is the capacity of mass disinformation – and that with “scientific authority” beyond what is possible by the social media today. In addition, we have to recognize, confirmed in several opinion polls, that because of decreasing trust in politicians, traditional media and scientists(!) there is more trust to be found in messages coming from “machines”. A survey run by the Collective Intelligence Project in 2025 found a consistent belief by a majority of people asked, that chatbots could make better decisions than their selected leaders.
AI promoting democracy in multiple ways
Some developers of and investors in AI became frightened because of this actual development and because of what may still come. Eric Schmidt and Andrew Sorota wrote recently an opinion in the New York Times – International Edition with the title “Use AI to reinvigorate democracy, not replace it!” They fear an “algocracy” developing instead of democracy. AI should be used to deliver better results to prevent that “people turn to strongmen, authoritarians and now algorithms, hoping for competence over chaos.” AI could map the landscape of public opinion, elaborate proposals that could bridge the people’s “divides rather than deepen them”. “It can make legislative texts accessible to ordinary citizens, explain trade-offs in plain language and help people articulate their preference. It can even scale deliberation across geographies and languages, making global coordination feasible in ways that were previously unimaginable.”
Well, it is good, that AI experts and gurus think also about how AI can help to build up local, regional and even international communities. But of course, politics must change itself independently from adaptation of AI to meet democratic demands. Politics and the whole educational system must try to counter the extreme individualization and try to come back to a discourse inside the different communities. And here politics could invite and use AI to help. In the words of Mark Coeckelbergh: “The idea of AI for democracy is to have AI foster and strengthen, rather than erode, democracy: to develop AI as a true communication technology in the republican and Enlightenment sense of the word. AI should then help us to realize that richer, relational ideal of democracy: it should help us to build a common world, develop common sense, and work towards the common good.”
In this respect AI would need filters which are preventing negative influences working against community building on the basis of serious exchange of ideas. They should be gatekeepers to protect democracy against anti-democratic forces and influences. Certainly, to put such filters and gatekeepers inside the different system of AI is not an easy task. It is technical difficult to prevent a so-called “jailbreaking”, which would subvert the filters with malicious command. And in addition, any filter could be interpreted as an instrument of censure-ship. It is not always easy to draw a clear and convincing line between filtering an AI to protect democracy on the one hand and censure-ship on the other hand.
One has also to mention the competition between different countries and ideologies which influence AIs. Can China and U.S. agree on the same principles and values which should be protected from misuse? And what is about Europe, when the present U.S. Vice President is expressing loudly his opinion, that Europe is neglecting free speech and when he is openly supporting right wing extremists in Europe. Maybe we can agree to prevent AI to give advice how to build a bomb, how to construct and distribute dangerous viruses, how terrorist could attack public institutions and facilities etc. But it will be much more difficult to agree what kind of filters should be applied to protect democracy.
Indeed, as Kissinger, Mundie and Schmidt write in their book: “The world needs a dedicated entity responsible for updating and refining the alignment’s training library, datasets, and validation tests.” And they add: “For their parts, government regulators should shape certain standards and audit models for assuring AIs compliance with them.” But they add also, that “enforcement of these standards could become extremely difficult.” The European Union tried and still tries to set such standards with its Artificial Intelligence Act. But as the Financial Times (November 20) writes under the headline “Can the EU fix its landmark AI rules”: “The fear-mongering about the existential risks of AI turned into a race for global dominance in the fast-developing technology – a race now defined by the power tussle between Washington and Beijing.” It is a difficult task to find the right balance between protection and precaution expressed in rules and incentives for curiosity and innovation. But it must be undertaken.
Sharing the extra profits
Another issue which is connected with democracy is the extreme wealth created by AI for those who are in the top league. Some fear that AI and the relevant companies are over-hyped and that could lead to a crash where many investors would lose a lot of money. Eric Schmidt, the former CEO of Google is of the opinion that AI is under-hyped. Whatever, for the moment and probably also in – near or distant – future many investors will earn an enormous amount of money. Parallel to the gains of a few, many could get unemployed and their families would suffer from the spread and use of AI. Therefore, some of the responsible AI experts think about how to share at least a small part of that increased wealth. In a discussion between Dr. Fei-Fei-Li and Eric Schmidt, the first said redistribution will be necessary, and Eric Schmidt underlined, that it is not possible to redistribute money in democracies(!). We need growth which will be created by AI and finally everybody will benefit.
But others are already thinking concretely about how to share the economic income or wealth created by AI. In a contribution of the NOEMA Magazine (November 22, 2025) its Editor-in-Chief Nathan Gardels presents two versions of sharing. The one proposal asks for a “tax the rich” method for redistribution of incomes. Aim is to create a Universal Basic Income (UBI) system. The other proposal asks to create wealth in the framework of a Universal Basic Capital (UBC). Interestingly, the latter is already included in a MAGA program – of course only for American citizens. Here the funding could also be done by distributing shares in the big tech companies. So, every U.S. citizen would become a capitalist and therefore – as the author underlined – such programs cannot be called socialist. The author concludes his analysis: “Adopting policies that foster universal basic capital for the AI era would place America’s off-track trajectory once again on the right side of history.”
Coming back to the start of this essay, the civilizational technology of AI cannot be treated as any other technology which is just having an impact of one or a few parts of our society. AI has and will have in future even more impact on nearly all sectors of our life on this earth and even beyond. Democracy is for many of us the best way how our community can be organized. It offers a mixture of individual freedoms and of mutual responsibilities. We have to be aware of the influences AI has on democracy and should choose the best way to adapt to this new technology.
Conclusion
As human beings we have different ways to adept to this new world. We can let the development go its way and submit ourselves to the new masters of the universe. Or we can try to be in a controlling position as long and as far as possible. In this case we have to construct the different AIs with filters, including filters which save the main values and principles we developed of many centuries. But we should also invite AI to play an active part in enhancing democracy. But in any way, we have to organize a sophisticated co-evolution of ourselves and the AI. As Kissinger et. al. write: “AI functions first as our main threat and then, ideally, as our partner.” Let us hope the human beings are managing such a process by accepting AI as a difficult, complex but also helpful partner. As Europeans we should be in the lead in implementing such a human being – AI partnership.
Recently David Gruber, a marine biologist who is also an AI expert and user, explained in a comment (New York Financial Times – International Edition November, 24th) how AI is helping to understand communication between sperm whales. He concludes his observation: “Used carelessly, AI could reinforce old hierarchies, treating both people and animals as data points. It could accelerate ecological harms under the guise of progress. But if used with humility and care, AI can become a bridge that reconnects us to the natural world…..To me this is the real potential promise of AI: not to make us faster or more efficient, but to make us wiser.” That wisdom is what we need to save democracy and make it functioning better in a multiple fractured world.
Autor: Hannes Swoboda ist Vorstandsvorsitzender des Sir-Peter-Ustinov-Instituts zur Erforschung und Bekämpfung von Vorurteilen und Präsident des International Institute for Peace.