The Ethics of AI Governance
The expansion of the social network and its consequential norms has undeniably impacted our modern societies. From a wave to a like, a share, and a retweet, the way that our demos communicates is unlike [anything] we have seen in previous decades. While these technological innovations have greatly advanced connectivity throughout domestic and global societies, it has also yielded many civil rights challenges for current government bodies. In the face of rapidly evolving social communication, we have experienced a complete metamorphosis in the way information is both shared and transmitted throughout the modern demos. The introduction of digital media, primarily social media, like Instagram, Twitter, and Tiktok, have proven to be an extremely valuable tool but at a hefty price. This leaves many experts fearful the current structure for governing information transmission and its subsequent protection is inadequate. As the current form is severely lacking the adequate legal and ethical frameworks to tackle the consequences of digital media.
Eva Erman, a professor and Head of the Political Science Department at Stockholm University, offered some insights into how the framework should be structured to accommodate the needs of our ever-changing relationship between artificial intelligence (AI) and government bodies around the globe. While some have argued that the only way to ensure an effective regulation of AI technology is through a single political consultative body, Erman contradicts the necessity of a singular governing body and proposes a more fluid, pluralistic arrangement tasked with protecting ethical boundaries. She contends that, “the discussion must include other [actors]…”, amongst which she points to AI developers, NGOS, philosophical consultants, and of course civil society. But she warns that relying on the ethical dilemma presented by AI interaction’s with governing structures must not be diluted to simply an ethical debate:
“...I think too much focus has been on ethical aspects of specific applications like self driving vehicles, autonomous weapons and so on. And I think that has… directed very little focus on how AI development is governed. So in that sense, I think much more is needed to include many actors when it comes to the governance of AI.”
The solution to the most crucial and daunting question facing our generation, of politically active and philosophically curious, must combine both policy implementation in congruence with ethical exploration of our socially defined value systems.
The 2021 raid on the US capital or the 2010 Arab Spring, amongst other failures, have indicated a clear ethical neglect and policy flaw to legally reign in digital reach. The underlying similarity amongst the above-mentioned failures for the global AI governance have cultivated a sentiment of skepticism and unwoven the social fabric of trust regarding federal political bodies and their constituents . The consequential polarity and disintegration between the people and their legislative leaders is Erman’s greatest worry . We must, she says,…
“...have governance mechanisms that focus on the societal goals and the means by which we achieve them. So in democratic societies, for instance, these are precisely the kinds of goals that we would want those who are affected by AI technology to be able to collectively influence. So in that sense, I think one of the most important procedural aspects here is democratic legitimacy in this kind of governance.”
The lack of collective trust that has come as a byproduct to AI proliferation has reignited one of the most historically relevant debates to exist: Who should have authority to regulate AI and to what extent should this authority stretch? It is obvious that those in the driving seat of the social media platforms and their subsequent algorithms ought to be held relatively accountable for their role in the relationship between the government and the demos. Yet, the multinational, corporate executives then must take on a role defined by juggling their profit driven business models all whilst protecting the integrity of society’s fabric. Unlike the previous decades’ political leaders, who still have had diplomatic faults, this agenda is involuntarily inherited by people who have no experience or concern for social stability. Although the reach of these organizations and their respective leaders is basically equal to that of global governments and their respective politicians, Erman continues,
“It's not self-evident that people who have influence should have influence just because they have a high capacity of influence.”
We have seen these leaders take what Erman dubs as a ‘soft law approach” to becoming an heir to the role of political responsibilities. Yet, she worries that crucial pillars of democratic rule such as “...transparency…, accountability, fairness, [and] participation” are frequently neglected and have taken a back seat to increases in profit and user engagement. Without a combination of ‘soft law,’ alterations in algorithms, re-figurations of administrative priorities, as well as ‘hard law’ legal frameworks, there will continue to be a persisting discontentment between global governance regimes and their constituencies within AI media, undermining these regimes’ political legitimacy.
Pauline Palama, a research assistant at the United Nations Educational, Scientific, and Cultural Organization (UNESCO), attributed the difficulty in fostering and protecting political legitimacy within the scope of AI to three factors: a lack of standardization for ethical assessments, a fundamental disparity amongst governing mentalities and cultural contexts, and finally the disconnect between politicians and computer scientists.
She highlighted the lack of understanding from both groups of experts: Politicians often lack the computer engineering and psychological knowledge to conceptualize the impact on behavior that result from social media algorithms and their inherent biases. By contrast, computer scientists generally lack the interdisciplinary education to ethically analyze their mathematical designs. Furthermore, within the cultural contexts that these algorithms take form, issues of cultural integrity protection often arise. Permeating across borders does not, however, constitute the greatest struggle for corralling the digital serpent of AI algorithms, Palma clarifies. Rather, the key problem of AI algorithms is their ability to contort and react within a millisecond, leading to significant behavior shifts. This becomes a huge complication for the success of global governance in the context of AI regulation as this process has historically taken months, years, possibly even decades to conform and evolve to the requests of the demos. While the reactive nature of the social media algorithms sounds like a democratic dream, it is deficient when it comes to proper philosophical and political deliberation. And without proper education for digital citizenship, moody dissatisfaction of the demos can take the driver’s seat when determining integral societal structures.
Therefore, unlike Erman, Palma advocates for an assortment of general, political, and expert groups that work both in opposition and in unison to maneuver through the many contexts, skill sets, and nature of each respective organization with its hands in the societal outcome of AI global governance. Notably, however, both experts, Erman and Palma, never spoke of a world in which AI no longer influences society or even governing structures. But she likewise clarifies the importance in harnessing its influence to further social democracy, as Palma explained:
“The thing is, I also think that this is not necessarily incompatible with the benefit for people..we also want to develop technology that allow human rights to flourish,...so we want to encourage …positive development of AI, [so that AI can] empower people [to] increase democracy potential.” (Pauline Palma)
Finally, Palma also discusses the complication of maintaining legitimacy within the bounds of the digital medium at a citizenship level, as it conflicts with the very procedural precedent for citizen input and legislative validity. In the other industries in which this relationship is evident we must employ a completely new philosophy of citizenship and participation, concerning the authentication of citizens’ identity. Whereas we would typically allow one political and social identity per human being in procedures such as social security, voting, or birth certificates – all of which act within a legal framework of certifying the singular identity of citizen, that philosophy is inapplicable in cyberspace. And this is fundamentally so, because, as Palma puts it :
“How do you ensure that somebody is who they say they are [online and then in real life]. And then that interfears on the very inherent nature of the digital realm, which is, allowing for the anonymity of you know someone [and] should only one account be allowed per person on Earth?”
Where must the legal regulations be imposed upon the beauty of anonymity that AI has given our modern social networks? Erman and Palma alike advocate strongly for the integration of an external regulating body that is both politically and philosophically analytical. With algorithms binging on hours of our attention, influencing our cultural and political identities, and allowing our social experience to permeate borders, we must question the ethical bounds of such a new and expansive tool. A tool which both researchers indicated they feel has much to offer for global governance and democracy. But the global standard for democratic rule is being put to the test by AI’s increased connectivity of the social network and we as a society must decide what the next step looks like. With this tool undoubtedly here to stay, we mustn't take a hands off approach but rather develop innovative archetypes for citizenship rights, political legitimacy, and ethical authenticity within cyberspace.
