Deciphering the Bletchley Declaration and White House Executive Order

Last week twenty-eight countries including the USA, China, Japan, and Germany attended an AI summit hosted by the UK culminating in the signature of the Bletchley Declaration: a resolution by attendees to ensure human-centric and responsible design, development, and exploitation of AI.

Why it matters

In what the UK government are calling a ‘world first’, several countries leading AI development signed a joint (if non-binding) declaration outlining a shared vision as to how to deal with AI safety risks and agreeing to set up an international network of AI researchers. The declaration includes a strategy to identify risks and build risk-based policies to ensure the safety of Frontier AI (highly capable general-purpose AI models that can perform a wide variety of tasks and match or exceed the capabilities present in today’s most advanced models). Further summits in France and South Korea mean this declaration is the first in a series of efforts to ensure sustained international cooperation.

Key Takeaway

A key point for industry to note is international agreement on the need for increased transparency by private actors developing Frontier AI. This reinforces the need for any businesses developing or using advanced AI tools to ensure that they assess, monitor, and document their use of AI to ensure they can demonstrate compliance with AI laws, policies, and standards as and when these come into force. A useful starting point for principles to consider is the voluntary Code of Conduct for AI developers under the Hiroshima AI process agreed by the G7 in advance of the summit.

White House Executive Order

Whilst the summit did not result in any binding obligations for nations or AI developers, two days before countries met, US President Biden signed an Executive Order on Safe, Secure and Trustworthy Artificial Intelligence which was shortly followed by the release of a draft policy on agency use of AI. With US companies leading development of Frontier AI, US lawmakers will play a key role in establishing any international frameworks for managing AI safety risks. Time will tell whether international collaboration will shape national AI approaches, or if leading AI nations will set de facto global standards through their own regulatory frameworks.

Notably, the US Executive Order explicitly addresses workplace disruption, a topic discussed by Prime Minister Rishi Sunak and Elon Musk during a fireside chat at the summit. It acknowledges that AI may undermine worker’s rights, worsen job quality, or result in harmful labour-force disruptions and emphasises the need to consider the views of workers and labour unions to ensure that benefits from technological innovation are universal. The Bletchley Declaration did not include reference to developing mechanisms for involving individuals in AI governance however, we expect that future summits are likely to discuss public involvement in defining AI policy.

Next Steps

In the King’s Speech on 7 November 2023, King Charles made no direct reference to general AI regulation and we do not expect to see a shift in the UK Government’s approach from that put forward in the AI White Paper (March 2023). We expect further AI guidance from UK regulators, international organisations, the US and EU in the coming months and will provide updates relevant to businesses developing, using, or considering adopting AI tools as and when this becomes available.