UNGPs lens to managing human rights risks from Generative AI - Forum on Business and Human Rights 2023
Broadcasting UN Entity
Day 2 (Room XXVI) - Applying the UNGPs lens to managing human rights risks from Generative AI - 12th United Nations Forum on Business and Human Rights 27 - 29 November 2023
Brief description of the session:
- This panel discussion session convened by the UN B-Tech Project ("B-Tech") at the UN Forum on Business and Human Rights will discuss how human rights risks associated with generative artificial intelligence ("generative AI") can be addressed through application of the UN Guiding Principles on Business and Human Rights ("UNGPs"). Serving as the official launch for a series of B-Tech papers examining the human rights dimensions of generative AI, this session will explore (1) how the development, deployment, and use of these systems may pose risks to human rights; and (2) how company and State action to address these risks should be informed by the UNGPs. The session will feature panelists representing the UN human rights office, generative AI developers, civil society, and other stakeholders.
- Generative AI, including chatbots relying on large language models such as ChatGPT, is capable of generating novel media content (including text, music, images and videos) at large scale in response to text prompts from users. Generative AI applications are not only being developed and used by businesses in the tech sector, but are also seeing significant uptake and deployment in a range of other industries, including law, healthcare and apparel. Appropriately, optimism about this technology's benefits has been accompanied by debate about its adverse human rights impacts. Information about the extent to which human rights due diligence has been conducted in the development and deployment of this technology is limited, and there has been little opportunity for learning across the tech industry about effective approaches to preventing and mitigating human rights risks stemming from advances in generative AI.
- There is thus an urgent need to explore what constitutes the appropriate scope and practice of business responsibility in relation to generative AI. Identifying appropriate responses to this question and building alignment across industry, civil society and standard setters about expectations should draw on international human rights standards. In particular, the expectations set out in the UNGPs can provide authoritative and widely accepted guidance. Using these global standards as the initial basis for unpacking the scope and nature of corporate responsibilities can also provide a common foundation for constructive and robust dialogue.
- To present a taxonomy of human rights risks linked to the development and use of generative AI by both tech and non-tech companies.
- To summarise current promising practices and gaps in generative AI risk mitigation.
- To discuss the implications of the UNGPs for business conduct by generative AI developers.
- To explore how the UNGPs should be integrated into governance approaches to minimising generative AI-related human rights risks.
- To propose next steps for a variety of actors to ensure that generative AI is developed and deployed responsibly.
- Peggy Hicks
- Mohamad Najem
- Alexandria Walden
- Mark Hodge
- Laura Schertel Mendes
- Preetam Maloor