Data Privacy Brasil’s contribution to the Thematic Deep Dive of Artificial Intelligence and other Emerging Technologies of the Global Digital Compact

In september 2020, the United Nations’ Secretary-General released the report “Our Common Agenda“, in which it proposes a Global Digital Compact (GDC) to be agreed at the Summit of the Future in September 2024 through a technology track involving all stakeholders: governments, the United Nations system, the private sector (including tech companies), civil society, grass-roots organizations, academia, and individuals, including youth. The GDC main goal is to “outline shared principles for an open, free and secure digital future for all“.

In this scenario, the President of the UN General Assembly appointed the Permanent Representatives of Rwanda and of Sweden as Co-facilitators to lead informal consultations with Member States and Stakeholders. These consultations (including remote participation) initiated in January 2023 and will last until June 2023. The process was separated in thematic deep-dives.

Data Privacy Brazil Research Association (DPBR) is part of the Southern Alliance for the Global Digital Compact, which comprises civil society organizations from Bolivia, India, Kenya and other African countries, as well as Brazil. We welcome this important process and have already participated in the themes of internet governance and data protection.

Last Thursday, May 25, the deep dive of Artificial Intelligence and other emerging technologies  took place, but even though we registered and prepared a speech, we didn’t have the chance to express our points at the meeting. As other civil society organizations already  pointed out, notable preference was given to the speech of Member States, UN agencies, and the private sector, at the expense of human rights civil society organizations, which prevented the speech that had been prepared by the DBPR and other civil society stakeholders.

With the aim of encouraging transparent and meaningful multistakeholder participation, the DPBR now publishes what would be its oral participation:

“Data Privacy Brasil’s contribution to the Thematic Deep Dive of Artificial Intelligence and other Emerging Technologies of the Global Digital Compact

United Nations, New York

May 25th, 2023

Dear co-facilitators and others present at this consultation. My name is Paula Guedes, I am a researcher at Data Privacy Brazil Research Association, and we are part of the Southern Alliance for the Global Digital Compact, which comprises civil society organizations from Bolivia, India, Kenya and other African countries, as well as Brazil. We are pleased to be part of this important process.

The Southern Alliance published its contribution for the Global Digital Compact in which we mention a list of core principles for AI regulation and key commitments, pledges and actions that may be done by important stakeholders. I will try to address some of them in my speech today.

In recent years, we have seen an explosion of ethical guides and principles being created, but without the necessary cogent force for their application in practice. Therefore, today we are talking about the need to transform ethical principles into effective practices by regulated agents, in order to enable a change in behavior that protects fundamental rights within the scope of AI. We are talking about binding regulation, which enforcement would be supervised by competent authorities, but through a responsive regulation framework, with participation of different stakeholders, not being monopolized by states.

In order to convert principles into practice, it is essential that the regulation provides governance instruments that are inspected and supervised by a competent authority with sanctioning powers. One of the outstanding instruments is the algorithmic impact assessment prior to the use of AI systems and with meaningful  public participation and due public scrutiny.

We need to overcome the idea of false tradeoffs that regulation would impede technological development, but think that regulation can serve as fuel for a more responsible technology that protects rights. In this sense, with the main goal to protect human rights, while promoting the technological development, the regulation should harmonize the risk-based and rights-based approach, in a way that AI systems are subject to more or less obligations and duties according to the level of risk, which means, that the regulatory intervention should be proportionate to the level of risk. Theremore, the greater the risk level, more obligations for the AI systems, corresponding to more rights to the individuals.  In this sense, through a precautionary approach, AI regulation should also establish uses of AI that are so harmful to fundamental rights that must be previously prohibited, as is the case of social scoring systems and facial recognition for public safety purposes.

It is crucial to have truly global awareness of the asymmetries between countries of the Global North and Global South and promote global best practices in AI to reduce these disparities, especially in terms of data collection and processing. At this point, regulation must consider the context in which it is applied,  always taking into account the prohibition of unfair discrimination against individuals or groups, especially the already marginalized and vulnerable population.

Thank you!”