Bias, Justice and Democracy: A Constitutional Perspective

Author:

Domenico Gabriele Nuzzolese - Lawyer & PhD Student in Constitutional Law at University of Bari Aldo Moro

Abstract

The relationship between bias, justice, and democratic processes has become significantly complex in contemporary democracies, particularly in light (and as a consequence) of digital transformations: algorithmic systems now mediate a large portion of our information ecosystem, posing serious questions about how the biases incorporated in these systems can impact democratic functioning and justice.

An analysis of the impact of algorithmic systems on the democratic processes represents a central element of contemporary constitutional and political research. In fact, the Literature highlights how Artificial Intelligence (AI) and the algorithms that constitute its operational foundation assume a relevance that exceeds the mere technical dimension, directly affecting the very premises of constitutional democracy as we have interpreted it to date (1).

In the digital era, algorithmic bias represents one of the most significant challenges faced by democratic societies because, as the decision-making processes are becoming more and more reliant on computational systems, the embedded prejudices of those algorithms could systemically disadvantage certain groups or distort the informational landscape.

Those biases are not merely technical flaws but reflect deeper structural inequalities and power dynamics.

Algorithmic Governance

Constitutional literature highlights that algorithmic biases emerge through multiple mechanisms: from distorted training data that reflect historical inequalities to design choices that prioritize certain values over others (2). For instance, content recommendation systems often prioritize engagement over informational quality, creating what we refer to as "filter bubbles" or "echo chambers", which divide the public sphere into separate informational environments (3).

On a Constitutional level, those implications are significant.

As recent constitutional studies reveal, the algorithmic profiling of information can be elevated from a mere technical-operational issue to a matter of constitutional relevance, directly related to the protection of informational pluralism (4), this because when algorithms determine which information is to be received by citizens, they fundamentally shape the conditions necessary for democratic deliberation.

In democratic societies, Justice with a capital 'J' requires equity, transparency and accountability in the exercise of power and the rise of what has been labelled as "algorithmic governance" challenges these fundamental principles in different ways (5).

Firstly, algorithmic systems often operate as a "black box", making their decision-making processes unclear not only to ordinary citizens but sometimes even to the developers themselves, leading to a lack of clarity that undermines the very principle of transparency, which is crucial for the legitimacy of democracy. It is important to remind ourselves that the core principle of Constitutional theory asserts that citizens must be able to understand how decisions affecting them are made in order to provide meaningful consent to governance.

Secondly, algorithmic systems could perpetrate and amplify existing social inequalities: research on "proxy discrimination" shows how algorithms can reproduce prejudices even when explicitly prohibited characteristics are excluded from their calculations (6). For example, elements such as postal code or digital behavioural patterns can serve as proxies for protected characteristics such as race or socioeconomic status.

Finally, the concentration of algorithmic power in the hands of Big Tech raises profound questions on the matter of accountability: a recent research noted that the privatization of control over information flows, marked by an unprecedented level of pervasiveness, necessitates the development of new tools to protect pluralism (7). When private entities determine the architecture of public discussion, traditional constitutional reviews of power may prove insufficient.

Right to Information: A Pillar of Digital Democracy

Democracy depends on the quality of information available to citizens and the fairness of the processes through which collective decisions are made, and both these dimensions are increasingly mediated by algorithmic systems nowadays.

What happened with Cambridge Analytica in 2018 exemplifies how algorithmic manipulation could undermine the democratic processes (8): by collecting personal data to create detailed psychological profiles of electors, political actors have gained an unprecedented ability in microtargeting messages, potentially manipulating electoral results. This case revealed how the convergence of big data, algorithmic profiling and political advertising creates new vulnerabilities in democratic systems.

Constitutional scholars pointed out that the right to information doesn't stop at the access to information, but also concerns its quality and pluralism (9), which adds a new qualitative dimension to the right to be informed, one that transcends simple content accessibility and requires the development of innovative legal instruments (10).

When algorithms filter content, they systematically prioritize or suppress information based on commercial logic rather than public interest, basically reshaping the informational environment on which democracy depends.

The European Union (EU) positioned as the leader in developing normative frameworks to face algorithmic biases and protect democratic values within the digital environment. Notably, I am here referring to The Digital Services Act (DSA), the Artificial Intelligence Act (AI Act) and the European Media Freedom Act (EMFA) which collectively represent a new approach to digital governance, prioritizing transparency, accountability and Fundamental Rights (11).

The AI Act carefully establishes systems used in democratic processes as "high risk," requiring additional safeguards and transparency measures (12) and, similarly, the DSA imposes obligations on very large online platforms to assess and mitigate systemic risks, including those related to electoral processes and civic discussion (13).

These regulatory innovations reflect the evolving understanding of constitutional principles in the Digital Era.

Conclusion

The relationship between bias, Justice and Democracy--capitalized here to emphasize their most noble essence aligned with Human Rights---in the Digital Era requires not only technical solutions but also constitutional innovations: since algorithmic systems increasingly mediate our information ecosystem and democratic processes, constitutional principles must evolve as well to ensure that these systems serve democratic values rather than undermine them (14).

This requires a recontextualization of Fundamental Rights, the development of new institutional mechanisms for algorithmic accountability, and the redefining of the lines between public and private power. As noted by recent studies, contemporary constitutionalism faces a fundamental challenge: the elaboration of new legal tools able to preserve the authenticity of public debate, whilst not suffocating technological innovation (15).

The challenge we face is not only technical, but political and constitutional in nature. It is up to us only to ensure that algorithmic systems function to strengthen our constitutional rights, rather than weaken them and render years of struggle futile.

References

(1) De Minco G., 2021. "Fundamental rights, European digital regulation and algorithmic challenge". MediaLaws.

(2) O'Neil C., 2017. "Weapons of Math Destruction: how big data increases inequality and threatens democracy". Washington: Penguin Books Ltd.

(3) Sustein C., 2017. "#Republic: Divided Democracy in the Age of Social Media". New Jersey: Princeton University Press.

(4) De Gregorio G., 2022. "Digital Constitutionalism in Europe. Reframing Rights and Powers in the Algorithmic Society". Cambridge: Cambridge University Press.

(5) Cardon D., 2016. "Che cosa sognano gli algoritmi. Le nostre vite al tempo dei big data". Milan: Mondadori.

(6) Zuddas P., 16 March 2020. "Intelligenza artificiale e discriminazioni". Liber Amicorum for Prof. Pasquale Costanzo - Diritto Costituzionale in Trasformazione.

(7) Betzu M., 2022. "I baroni del digitale". Naples: Editoriale scientifica.

(8) Kaiser B., 2019. "Targeted: The Cambridge Analytica Whistleblower's Inside Story of How Big Data, Trump, and Facebook Broke Democracy and How It Can Happen Again". New York: Harper Collins.

(9) Papa A., 2019. "La tutela del pluralismo informativo nell'ecosistema digitale". De Iustitia.

(10) Pitruzzella G., Pollicino O., 2020. "Disinformation and Hate Speech: A European Constitutional Perspective". Milan: Bocconi University Press.

(11) Caggiano G., 2023. "Il quadro normativo del mercato unico digitale". In F.Rossi Dal Pozzo (Ed.), "Mercato unico digitale, dati personali e diritti fondamentali". Eurojus.

(12) Tuozzo M., 2024. "Note intorno al (futuro) Artificial Intelligence Act. Prospettive costituzionali e sfide dell'immigrazione nell'ecosistema digitale". Diritto Pubblico Europeo Rassegna Online.

(13) Orofino M., 2024. "Il Digital Service Act tra continuità (solo apparente) ed innovazione" in Pizzetti F. (Ed.), "La regolazione europea della società digitale". Turin: Giappichelli.

(14) D'Aloia A., 2020. "Il diritto verso il mondo nuovo. Le sfide dell'Intelligenza Artificiale in Intelligenza artificiale e diritto" in D'Aloia A. (Ed.), "Intelligenza artificiale e diritto". Milan: Franco Angeli.

(15) Simoncini A., 2019. "Il cambio di paradigma nell'intelligenza artificiale e il suo impatto sul diritto costituzionale". Rivista di filosofia del diritto.

Biography of the Guest Expert

Domenico is an administrative Lawyer with expertise in public procurement and concessions, with a particular focus on renewable energy issues. He practices law at a prominent firm in Puglia and simultaneously conducts research as a PhD student in Constitutional Law at the University of Aldo Moro in Bari (Italy).

Domenico’s academic background includes a Law degree from the University of Aldo Moro in Bari (Italy), with a thesis focused on electoral systems. He later enhanced his studies through the Advanced Training Course in Constitutional Justice and Judicial Protection of Rights at the University of Pisa (Italy). He is currently pursuing a PhD at the University of Aldo Moro in Bari, centered on the topic: “The Evolution and Protection of the Right to Information in the Digital Era”.

Domenico actively contributes to the scientific debate on topics such as bioethics and AI by publishing articles in specialized journals and participating as a speaker in nationally and internationally significant conferences.

His intellectual background is enriched by an long musical journey at the Conservatory of Music in Matera, an experience that has provided him with a broader, multidisciplinary cultural perspective, integrating rigorous legal training with artistic sensitivity.