As artificial intelligence (AI) rapidly transforms how our societies operate, it also affects our human rights, presenting both opportunities and challenges. They include, among others, concerns about privacy, discrimination, information integrity, and widening inequalities. To better understand AI’s impact on human rights, UNRIC spoke with Iverna McGowan*, Tech & Human Rights Advisor at the UN Office of the High Commissioner for Human Rights. Iverna explains how AI can exacerbate existing risks, how it can help strengthen human rights protection, and how we can ensure everyone has a voice in shaping the future of AI.
What are the main human rights implications of the rapid growth of generative AI?
Generative AI can be used in many ways, and looking at this question from a human rights perspective is a good start. Human rights give companies, regulators, and civil society a clear framework for identifying and addressing the impacts of generative AI systems.
It is important to recall that the implications depend on the type of AI tool, the sector concerned and the broader context, for instance, the state of the rule of law in a given jurisdiction.
At UN Human Rights (OHCHR), through our B-Tech project, which engages with the private sector in particular, we have developed a taxonomy of human rights risks associated to generative AI. This paper has been used by various actors working on the implementation of the ‘EU’s AI Act. The paper highlights how generative AI can put human rights at risk. Many of these risks existed with earlier types of AI, but generative AI can change them, increase them, or create novel issues.
One concrete example is privacy. Generative AI models might be trained on data that includes personal or sensitive information. Because these models often use large datasets scraped from the internet, people may not know—or be able to give informed consent—when their data is collected, used, or stored for AI training. So, whilst the issue of consent and personal data is not new, generative AI can bring this challenge to a new scale.
How can AI deepen existing inequalities and disproportionately affect marginalized communities?
Discrimination and inequality are not marginal issues; they are systemic failures that compromise the integrity of societies, democratic, participatory governance, and the realization of human rights for all. We must urgently act to ensure that digital transformation does not turbocharge historic discrimination by embedding it in AI models. This issue of discrimination and AI was the topic of a new OHCHR report.
We have already seen many examples of how AI can deepen existing inequalities. In the context of law enforcement, predictive policing and biometric recognition systems can disproportionately affect minorities, reinforcing historical injustices. In social protection, automated eligibility systems can exclude at-risk or vulnerable groups, increase surveillance, and lead to inaccurate decisions with serious consequences.
A key challenge with AI is transparency. If someone does not know that a decision of consequence was made about them using AI, then they may not think to, or know how to challenge that decision. This also has human rights implications, as it negatively impacts people’s right to an effective remedy.
What can be done to prevent algorithms from amplifying hate speech and disinformation online?
One important but often overlooked issue is how personal data is used. OHCHR recently published a briefing paper on the topic of Platform Governance and Human Rights, that summarizes some of the key guidance from the international human rights system in this area.
To target advertising and maximize user engagement, most platforms decide whether to use algorithms that amplify or suppress content based on data they collect about users, e.g. regarding ‘people’s preferences, demographic and behavioral patterns. This targeting can fragment the information ecosystem, increase social polarization and erode information integrity. Targeted advertising and profiling also drive and exacerbate the spread of disinformation, threatening both privacy and informed discourse.
We must prevent the use of private and sensitive data for targeted advertising without consent, particularly when it affects groups such as children. A human rights-based approach involves focusing more on process and procedure rather than on the problematic practice of labelling individual pieces of content as true or false. This is why OHCHR stresses the importance of transparency, including ad-libraries and mandatory transparency reports, and that platform governance efforts be firmly anchored in the right to privacy.
Can AI strengthen information integrity and improve access to reliable news?
AI can help strengthen information integrity, but only if it is well-governed and anchored in human rights. AI alone cannot guarantee reliable news. An enabling environment is key, including investment in independent media and protection of journalists.
AI systems can be used for content moderation, but they still struggle with context and nuance. This is why transparency and human oversight remain so essential.
How can AI reinforce the protection of human rights? Where do you see the biggest opportunities?
AI can be a powerful force for public good — unlocking economic growth, and enabling advancements in health care, education and beyond. Yet, to ensure that AI serves the public interest, we need a deep understanding of human rights impacts and must allow for restrictions, including prohibitions on technology when sufficient guardrails are not in place.
Ensuring participation, transparency, accountability and non-discrimination is key. These principles make AI systems trustworthy and human-centric, and ensure that innovation responds to ‘people’s needs at the local level.
What role is the UN Human Rights Office playing in shaping global standards and guidance on AI governance?
The UN Human Rights Office plays an important role in shaping global standards for AI governance. Under the UN Global Digital Compact, through the advisory service on human rights and digital technologies, it provides advice and guidance on digital technologies to a wide range of stakeholders. Many States and AI safety institutes examine how useful international human rights law is as a preexisting and authoritative framework for navigating novel questions about how to ensure AI serves humanity.
Recently, we were called to testify before the Brazilian parliament as Brazil debates its AI Act. UN Human Rights has also developed 10 recommendations to help States regulate AI at the national level.
At the international level, the UN Human Rights Office works closely alongside other UN entities to support the establishment of the Independent International Scientific Panel on AI and the global dialogue on AI governance. Human rights are central in these international efforts. Our Office will be engaging on crucial questions such as meaningful civil society participation and ensuring that core human rights principles and considerations are at the forefront of deliberations. The Office also works on technical standard setting and human rights.
How can we ensure that everyone has a voice in shaping an AI-powered future?
At UN Human Rights, we firmly believe that participation in public affairs is not only a human right, but it also strengthens outcomes and avoids harm. Meaningful public participation should be a basic requirement in AI regulatory processes. States must respect their obligations to uphold human rights, they must lead participatory processes and ensure that the public has the necessary tools, including transparency and avenues for remedy in cases where AI systems are being used to make decisions or provide services.
Also, companies have a clear responsibility to ensure meaningful stakeholder engagement, as per the UN Guiding Principles on Business and Human Rights.
*Iverna McGowan is currently serving as Tech & Human Rights Advisor at UN Human Rights (OHCHR). She has been listed by EU Insider in 2024 as one of the most influential people on digital policy and by POLITICO in 2017 as one of the most influential women on EU policy and law making. Iverna was Executive Director of Amnesty ‘International’s Office in Brussels and Advocacy Director for Europe for seven years and covered EU and Council of Europe issues. Prior to that, she worked in the Human Rights Unit at the Irish Department of Foreign Affairs.
Further reading
Artificial intelligence – United Nations Western Europe
Building an accessible future for all: AI and the inclusion of Persons with Disabilities
From Braille to AI: Shaping a more accessible future
UNRIC Library Backgrounder: Artificial Intelligence – Selected Online Resources