14 November 2023
By Pablo Fuentes Nettel
In the past few weeks, the field of AI governance has witnessed some major developments with events like the AI Safety Summit, the G7’s statement on the Hiroshima AI Process, and the White House’s release of the Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence. These events have definitely put AI in the spotlight, but that’s just the tip of the iceberg. As someone who works on AI policy, I keep stumbling upon events on AI governance all the time —summits, conferences, keynotes, and seminars seem to be happening everywhere. As much as I enjoy reading and listening to my colleagues, I can’t help but wonder whether this AI Summit Boom is making a difference in how AI is governed. So, to organise my thoughts, I wrote down a couple of initial reflections about this phenomenon.
In the first place, it seems like we are witnessing a scenario of normative fragmentation. There are many organisations and thought leaders telling other actors what they should do. We see events organised by governments, tech companies, international organisations, non-profits, think tanks, and the media. They all have unique perspectives and interests. Hence, they address discussions differently, starting from varied questions and using diverse definitions. There’s no doubt that diversity is positive; but when discussions are siloed, reaching a consensus to create cohesive policy frameworks becomes more complex.
On top of that, the global scope of AI governance adds another layer of complexity. When we talk about AI governance there’s no one-size fits all solution. By talking to government officials for the development of OI’s Spotlight Series, we’ve had the opportunity to learn about multiple approaches to AI governance. Governments address AI in various ways, with notable differences in how they deal with it and in their perceptions of risks and opportunities. The current global scenario is one of differing economics, legal systems, and political dynamics, as well as varying levels of technological readiness.
This comes with important risks. This diversity of perspectives may potentially contribute to discussions that appear somewhat superficial, as if some of these events serve primarily as symbolic gestures —perhaps, hosting AI gatherings has become a trend without really delving into the complexities of AI governance. Additionally, there is a concern that, in the middle of this messy landscape, the more influential players may assume a disproportionate role in defining guidelines, adopting a predominantly top-down approach. In that sense, it is imperative to highlight that a robust governance framework demands the comprehensive and inclusive participation of all stakeholders.
Navigating the AI Summit Boom, hence, can be overwhelming, often leaving us sceptical about the state of AI governance. The sheer chaos in discussions and diverse perspectives might make it seem like finding common ground is an impossible mission. Yet, in the midst of this diversity, we can’t forget a critical element for any international governance framework: acknowledging diversity and setting standards aren’t mutually exclusive. Yes, it is possible to embrace the mess brought by the AI Summit Boom to come up with differentiated governance mechanisms while identifying common ground among states. While it’s easy to get lost in the AI buzz, there are noteworthy developments that show promise. Notably, there is a global trend towards adopting the OECD’s AI Principles, and governments are progressing into the implementation phase of UNESCO’s Recommendation on the Ethics of Artificial Intelligence. Similarly, at Oxford Insights we just launched the Trustworthy AI Assessment Tool, which aims is designed to help policymakers in government understand how prepared their government is to use AI in a trustworthy way in the public sector. These initiatives acknowledge that the global AI landscape is diverse and involves different levels of AI readiness, while setting minimal standards that contribute to a more robust governance system.
Beyond the messy nature of the AI Summit Boom, there’s a pretty evident advantage: the governance of AI has become a hot topic. As the discussion gains traction, people more aware of the societal implications of AI developments. Increasing awareness brings a range of benefits. First off, putting AI in the spotlight on the political agenda is a big plus. The increased public interest in AI, prompts governments to allocate more attention to it, potentially leading to higher levels of commitment, accountability, and transparency.
Similarly, more awareness opens up additional spaces for dialogue and exchange of best practices. Each day, additional governments join the discussion, sharing their strategies for addressing AI challenges and offering insights into effective approaches. The global community is actively participating in this conversation, sharing experiences to illuminate both successful and unsuccessful endeavours. This collaborative element can serve as a catalyst for crafting comprehensive and inclusive policies and governance frameworks.
Additionally, more awareness can translate into increased funding for AI governance projects. As both public and private stakeholders recognise the relevance of AI, they channel more resources into research endeavours and the establishment of effective regulations. Securing funding is pivotal for ensuring the sustained viability of AI governance initiatives in the medium and long term.
Finally, another crucial benefit has to do with AI literacy. As individuals become more informed about the challenges and opportunities presented by this technology, societies are better equipped to make informed decisions. They can actively participate in shaping regulations for AI that are equitable and sensible for all. This is essential in addressing the risks associated with AI, ensuring that our regulations are comprehensive and protect everyone’s interests.
With all its complexities, the AI Summit Boom has undoubtedly put the spotlight on the evolving landscape of artificial intelligence governance. Despite the challenges posed by normative fragmentation and diverse perspectives, positive strides are evident through initiatives led by organisations like the OECD and UNESCO.
Navigating the AI Summit Boom might be overwhelming, but embracing this complexity allows for the development of nuanced governance mechanisms while still identifying common ground among nations. Moreover, the increased awareness generated by the AI Summit Boom can be a driving force for public sector reform, prompting governments to double efforts to harness this technology for the public good.
As we look ahead, predicting the trajectory of AI governance remains uncertain. Perhaps, a valuable tool would be delving into the historical trajectory of other major global issues, such as climate change and nuclear technologies. Comparative analyses might suggest that the path forward for AI governance will likely involve a transition from initial chaos and uncertainty to more structured, collaborative approaches. Simultaneously, it could shed light on the potential asymmetries that could be institutionalised if we don’t apply an inclusive approach. Anyway, this is a very complex topic that requires further exploration in a subsequent article.
P.d. At Oxford Insights we just launched the Trustworthy AI Assessment Tool, which is designed to help policymakers in government understand how prepared their government is to use AI in a trustworthy way in the public sector. Check it out!