Professor Yuval Shany, Institute for Ethics in AI
Human rights are increasingly under threat as technologies like artificial intelligence (AI) evolve and integrate into our societies. Practices such as profiling and mass surveillance have raised significant concerns about equality and privacy. This shift invites us to reconsider how human rights protections can be expanded to encompass new needs and interests that emerging AI applications create.
AI holds remarkable potential to revolutionize sectors such as health, education, and work. However, it also poses risks that can lead to serious personal harm, unintended manipulations, and opaque social control mechanisms. The urgency is palpable—should our legal frameworks evolve at the same pace as technological advancements? Currently, our human rights laws appear outdated, lacking the necessary adaptability to protect citizens in an AI-driven future.
The spotlight has thus far been on safety, ethics, and human rights within the AI discourse, yet a comprehensive reassessment of human rights laws is imperative. Through collaboration with leading human rights centers across four continents, an intriguing solution has been proposed: an international AI Bill of Human Rights. But why is this idea gaining traction?
A New Mandate for Human Rights
Legal tools crafted in recent years by organizations such as the UN, EU, and various national governments aimed to fill existing gaps in the regulatory landscape for new technologies. Yet, a deep dive into these frameworks reveals shortcomings—often they provide only partial coverage or utilize vague language that fails to resonate with established human rights principles.
“We are rushing headlong into an AI-driven future, but our legal protections…remain stuck in the past.”
These deficiencies result in a fractured landscape of standards that struggle to effectively address cross-border AI systems. The lack of cohesive protections complicates the assignment of responsibility, especially for private companies when harm occurs. Hence, the proposition for a global solution emerges strongly.
Charting the Course for an International Bill of Human Rights
As a Fellow of the Accelerator Fellowship Programme at the University of Oxford, extensive research into the viability of an international AI Bill of Human Rights has revealed a pressing need for uniform standards. This initiative is bolstered by consultations with experts from leading human rights institutions in cities such as Oxford, Geneva, Pretoria, and Harvard.
The groundwork laid by these discussions has culminated in a concise Bill of Rights, outlining fundamental protections that individuals should expect wherever AI is devised or deployed. These protections align with reasonable public expectations, emphasizing innovation while safeguarding fundamental human rights.
“They codify what people reasonably expect: to benefit from innovation without surrendering their freedom, equality and dignity.”
Seven Fundamental Rights for an AI-Driven Future
In the White Paper on the Feasibility of an International AI Bill of Rights, I have proposed an initial framework comprising seven core rights:
- Access to AI: Ensuring access to safe and reliable AI tools and related technologies.
- Privacy protections against harmful uses of AI: Safeguarding individuals from invasive monitoring and data collection.
- Freedom from algorithmic bias and unfairness: Designing systems that prevent discrimination and uphold equity.
- Transparency and explainability: Providing clarity on how AI influences significant decisions affecting rights.
- Protection from algorithmic manipulation: Preventing deceitful practices that undermine autonomy.
- Human decision and human interaction: Allowing individuals to opt for human oversight in critical situations.
- Accountability for harms caused by AI: Establishing responsibility and pathways for remedy when AI inflicts harm.
These rights seek to promote ethical AI development and use, aligning technological possibilities with human welfare in ways that emphasize justice and fairness.
“The adoption of an international AI bill of human rights would support public confidence…and reduce legal uncertainty.”
The Imperative of Timeliness
The urgency for action cannot be overstated. The legal measures we implement today may be more feasible and cost-effective than retrofitting tomorrow, leading to critical gaps in protection as technology progresses. The intent is not to hinder AI advancements but rather to ensure a concordant evolution of both human rights and technological innovation.
This is a proactive approach to create systems that protect individual freedoms while harnessing the benefits of AI developments. To learn more about this pressing initiative, I invite you to explore the resources available in the White Paper, as well as tune into the enlightening discussion on the “AI and Human Rights” podcast produced by the Institute for Ethics in AI.