Can Europe force its new AI law on the rest of the world?
Undoubtedly, the European Union -even with its chiaroscuro- represents the contemporary world bastion of Democracy and, by extension, the champion of the fundamental rights of citizens as free people, by rigorous political and social comparison with the rest of the human orb. So much so, that only in ancient Europe does the political organizational system that we know as the Social Welfare State exist, with a deeply rooted humanist philosophy, of which I am a fervent defender as it is a key public management instrument for the good human development of the daily life of people. In fact, the socio-political models of Democracy and Social Welfare are neither universal values nor are they a logical evolutionary consequence of developed societies, as I have commented so many times (1), to the annoyance of the self-satisfied European ethnocentrics.
Within this contextual framework, the European Union has just taken another step, last Thursday, May 11, in defense of the guiding principles of Democracy with the provisional approval -by its Committee on the Internal Market and Civil Liberties — of the first global law on Artificial Intelligence (AI) (2). Law that must be ratified by the Community Parliament, foreseeably without relevant surprises, in the middle of next June. A legal norm with which the European Union intends to be a world leader in the orientation towards the development of a reliable, safe and human-centered AI, by regulating mainly in terms of prohibitions of both surveillance and biometric categorization, as well as recognition of emotions, and also of predictive surveillance AI systems; as well as in the regulation of tailor-made regimes for AI for general use and basic models such as GPT, with the aim of ensuring the transparency of knowledge and public information; and the implementation of the citizen’s right to present complaints about AI systems, under the protection of a future public authority for democratic control of AI. A regulatory aspiration that, it should be noted, will make illegal practices in European territory a large part of the current practices already socially normalized by the two great world powers in AI development: the USA and China, and that at the same time they will collide head-on with the current business policies of the leading technology companies in AI in the Market.
Against this background, it seems clear that the European Union is exercising -perhaps without being aware of it- the role of the new Athens of the 21st century, and that in its iron defense of democratic values it goes straight into the attack with the rest of the developed world in AI that plays the role of an emerging Sparta. An unequal confrontation that, I am afraid, will end up replicating the history of the Peloponnesian War of the s. V BC, where after a timid first period of initial peace between the City-States of Ancient Greece, the Spartans finally decided to defeat the Athenians, reducing them to a relationship of submission that, as a side effect, plunged them into a state of poverty of that they could never recover their former prosperity.
I must aim, as a European and a democrat by idiosyncrasy, my praise for the voluntarism of the European Commission in terms of AI control, also considering the processing of the new regulation as an obligatory political diligence and an inalienable humanist moral duty in present tenses. However, we must not lose objectivity in the face of the irrefutable fact of the scant political weight that Europe has in the world -to do the contrary would be to err on the side of naivety-, since although we represent 14 percent of the international trade in consumer goods (being the last in the AI development ranking, as an explanatory note), the new international geopolitical scenario that emerged from the Russo-Ukrainian war highlights our commercial (3) and consequently diplomatic limitations, which makes it impossible for the Union Europe to apply as a future superpower neither in the medium nor in the long term (4). Or, in other words, Europeans -in the past attached to the imperial dream- today represent a modest socio-political oasis of classical heritage, at the mercy of compression of the forces of tension typical of the powers that polarize the world.
In this scenario, the question is no longer so much: what force majeure measures are we going to deploy before the outside world to reaffirm our regulatory leadership? Since the European imposition capacity at the international level is known to all to be limited, but rather, can we guarantee the Local regulatory resistance to the prevailing external pressure of an AI developed in distant latitudes and under conflicting ethical parameters?. Faced with this last question, there are only two possible results -similar to the reduced options that luck offers us when a coin flies- with their respective consequences:
1.-If positive, the internal regulatory restrictions imposed on third parties on the development of AI to guarantee the health of the European democratic model can lead us, as is beginning to be glimpsed in the case of some AI consumables in light of the recent events (remember that Google has just left Europe out of its new AI, for the moment, due to the problems caused by the violation of the privacy of ChatGPT with respect to the community regulation on data protection) (5), to the fact that the large technology companies can leave operating in our Market in a partial or total way by their own will or imposed. A future assumption that would lead the old continent to a technological impoverishment and by economic extension in relation to the rest of the developed digital world, on the reliable basis of our global competitive incapacity in AI.
2.-While in the negative answer, the presumable lack of capacity to impose third parties on the development of AI to guarantee the health of the European democratic model would also lead us to a presumable state of technological submission that, on the one hand, it would end up irreparably devaluing our democratic standards and, on the other hand, it would increase our economic dependence on external technology providers at the gates of the Fifth Era of the Industrial Revolution.
Two statistically imminent scenarios in which, in any case, it does not seem that the euro zone market is going to be very well off. That is, everything points to the fact that the history of Athens versus Sparta is destined to repeat itself.
However, Europe could have to play a third way in its double attempt to impose its regulation on AI internally and, in turn, be able to hoist the new law as a humanist flame to enlighten the world, which is none other than to achieve the affiliation to his cause by a natural ally as a world power, with the USA being the only viable option. Which is remotely feasible, if not impossible, since the commercial interests of both markets intersect at a point of continuous conflict (in fact, the USA is economically interested in a weak and dependent Europe, as well as reveals its aggressive foreign trade policies and its impermeable domestic trade policies), without mentioning that the European humanist constitutional principles squeak with the ultraliberal North American principles, even when dealing with two socio-political spaces self-defined as democratic. Differences that, in addition, are currently aggravated in terms of development and social implementation of AI, especially in data protection (citizen privacy), biometric categorization (negative discrimination based on socio-cultural parameters) (6), and predictive surveillance. (based on profiles, location or past criminal conduct) (7), among many other fields of disagreements in roboethics.
For a contemporary Europe, a recreation of the new Athens of our century, whose society is based on its own democratic model that promotes social justice, collaborative commerce and humanistic knowledge, it really has little capacity for action against the rest of the world developed exterior with a Spartan spirit, characterized by a leadership based on economic or ideological ruling classes, a fiercely competitive trade and hungry for colonization, and a strong military disposition. Therefore, in this international reference framework, the proposal for a new European law on the regulation of AI, much to my regret, seems more like an illusory declaration of good intentions, the result of distinguished humanists, than a feasible regulatory measure in middle of an ethically dataistic AI world (8).
Disturbingly, after the news of European AI regulation has spread beyond the old continent, we can already hear the Spartan war cry (Au, au, au!) as a distant echo. And although the wind is blowing confrontational airs in terms of AI, and knowing that we, the well-to-do Europeans, will be able to oppose little resistance, it would certainly be a betrayal of our own collective legacy not to try to regulate AI under the strict control of the social values of our ancient culture. Even if the announced defeat goes with it. Well, after all, Ethics, and specifically in these times Roboethics, define what we are as a civilized society.
References
(1) Democracy does not exist as a universal value and it must be defended against external enemies (Notice to Euronavegantes). Jesús A. Mármol. A Seeker’s Log, October 17, 2021 https://acortar.link/8tl9ia
(2) AI Act: a step closer to the first rules on Artificial Intelligence. European Parliament, May 11, 2023 https://acortar.link/kxOscA
(3) Which countries are leading the artificial intelligence race? InvestGlass, February 6, 2023 https://acortar.link/F90GOp
(4) The EU in the world after the war in Ukraine. Federico Steinberg, Jorge Tamames. Elcano Royal Institute, April 21, 2022 https://acortar.link/trgVyQ
(5) Google leaves its artificial intelligence out of Spain due to the privacy problems of ChatGPT. Rodrigo Alonso. The World, May 11, 2023 https://acortar.link/utfPpx
(6) The Robot-Judges bring an era of fair but inhuman justice (due to lack of empathy). Jesús A. Mármol. Medium, October 28, 2022 https://acortar.link/iSkTKu
(7) Ethical dilemmas of AI that predicts crimes before they happen. Jesús A. Mármol. Medium, October 21, 2022 https://acortar.link/SFZRyD
(8) Dataism and Humanism, an incompatible relationship?. Jesús A. Mármol. A Seeker’s Log, March 10, 2020 https://acortar.link/EubfVc