New Report: AI Foundation Models & Detection Mechanisms

How can we reduce the risks stemming from general-purpose AI foundation models? In this 2023 report, GPAI Experts propose the establishment of a reliable detection mechanism for the content produced by AI foundation models, allowing consumers to determine whether an item was generated (wholly or partly) by AI.

Read the full report

GPAI Summit 2023

New Delhi, India

Ministerial Council: 13 December 2023
Full event: 12- 14 December 2023  

Global Challenge to Build Trust in the Age of Generative AI

Call for Partners

Recognising the transformative and disruptive potential of generative AI, the OECD, GPAI, and UNESCO are joining forces with AI Commons, IEEE, and VDE to advance global collaboration, trust, and transparency with regard to generative AI

Complete our Global Challenge Partner Inquiry form and get in touch ai@oecd.org

G7 Ministers Reaffirm Commitment to Advancing Generative AI Policy Through Multilateral Collaboration, including GPAI

September 7, 2023, Hiroshima

On 7 September, G7 Digital & Tech Ministers met within the Hiroshima Artificial Intelligence (AI) Process, in a stride to foster international cooperation in the realm of AI. They issued a statement emphasizing their dedication to collaborative initiatives on generative AI and common policy priorities.

This commitment is further fortified by the ministers' intent to collaborate with prominent international organizations and actors, among them the Global Partnership on Artificial Intelligence (GPAI). GPAI, at the vanguard of global AI initiatives, is ready to join forces with these distinguished leaders in the pursuit of policy discussions deeply rooted in empirical insights.

GPAI eagerly anticipates the continuation of its longstanding partnership with the OECD and other IGOs. This collaboration, exemplified by initiatives like the Global Challenge to Build Trust in the Age of Generative AI, jointly launched by GPAI and OECD.AI, builds on harnessing AI's boundless possibilities while ensuring ethical, responsible, and trustworthy implementation.