Nations Unite at Bletchley Park to Establish Historic Agreement on AI Safety

In a groundbreaking moment, leading AI nations, including the United States, China, and the European Union, have come together under the UK’s leadership to forge a historic agreement at Bletchley Park on November 1, 2023. This unprecedented accord, known as the Bletchley Declaration on AI safety, signifies a shared recognition of the opportunities and perils presented by cutting-edge AI and the imperative for collaborative governmental efforts to address the most significant challenges.

The Bletchley Declaration is endorsed by 28 countries spanning the globe, encompassing regions like Africa, the Middle East, and Asia, in addition to the European Union. It underscores the pressing need to comprehend and collectively manage potential AI-related risks through a new global initiative aimed at ensuring the responsible development and deployment of AI for the greater global good.

Countries supporting the Declaration include Brazil, France, India, Ireland, Japan, Kenya, the Kingdom of Saudi Arabia, Nigeria, and the United Arab Emirates.

This historic Declaration not only fulfills critical summit objectives in establishing a shared consensus on the risks and opportunities associated with frontier AI but also lays the groundwork for international collaboration on AI safety and research, with a particular emphasis on enhanced scientific cooperation. The discussions held at the Summit, involving leading frontier AI companies and experts from academia and civil society, further delve into understanding frontier AI risks and improving AI safety.

The participating nations acknowledge the substantial risks stemming from both intentional misuse and unintended control issues related to frontier AI. These concerns are amplified by cybersecurity, biotechnology, and disinformation risks. The Declaration highlights the potential for significant harm, whether deliberate or unintentional, arising from the capabilities of these AI models. It also recognizes risks beyond frontier AI, such as bias and privacy.

In recognition of the need to deepen the understanding of AI risks and capabilities that remain not fully comprehended, the participants have committed to collaboratively support a network of scientific research on frontier AI safety. This builds upon the UK Prime Minister’s recent announcement to establish the world’s first AI Safety Institute and complements existing international efforts within organizations like the G7, OECD, Council of Europe, United Nations, and the Global Partnership on AI. This collaborative approach aims to create a foundation of scientific research for managing AI-related risks while unlocking the technology’s benefits.

The Declaration underscores that the risks are best addressed through international cooperation. As part of the agreement for international collaboration on AI safety, the Republic of Korea will co-host a mini virtual summit on AI within the next six months, with France hosting the subsequent in-person Summit a year from now. More details about these events will be provided in due course.

The Bletchley Declaration, building on the UK’s recent initiatives in AI safety, also recognizes the responsibility of those developing powerful and potentially hazardous frontier AI capabilities to ensure their safety, including the implementation of testing systems and other appropriate measures.

UK Prime Minister Rishi Sunak expressed his optimism about this historic achievement, emphasizing the shared responsibility to address AI risks and promote international collaboration on AI safety and research. The UK, once again, takes the lead in pioneering discussions to ensure AI’s safety and its enduring benefits for future generations.

Technology Secretary Michelle Donelan highlighted the significance of this agreement in building public trust and ensuring the safe development of AI. She emphasized the long road ahead and the Summit’s role in initiating a lasting process to unlock the limitless potential of AI.

Foreign Secretary James Cleverly stressed that AI transcends borders and its impact on the world is deepening. He expressed the UK’s pride in initiating the global discussion at Bletchley Park on utilizing AI as a force for good.

The French Presidency welcomed the collaborative work initiated by the AI Safety Summit and pledged its active participation in the collective effort. France is ready to contribute to this long-term endeavor and will host the next in-person Summit.

The Republic of Korea Government, pleased to co-host the mini virtual summit, recognizes the importance of multilateral cooperation in ensuring the safe and responsible use of AI technologies.

His Majesty The King delivered a virtual address, acknowledging AI’s transformative potential and emphasizing the need for international coordination to ensure AI’s safety and security. He thanked the attendees for their role in establishing a lasting consensus on AI safety.

This historic Declaration represents a significant step in addressing AI risks on a global scale and serves as a testament to the collaborative efforts of nations to harness the potential of AI while safeguarding against its risks.