A focus shift from AI safety to AI opportunity at the 2025 AI Action Summit

By Cascade Senior Advisor Laura Mahrenbach

AI Action Summit

Paris, France | February 10-11, 2025

Photo by Mathias Reding on Unsplash

On February 10-11, 2025, the governments of France and India hosted an AI Action Summit in Paris. The summit was attended by over 1000 government, business, and civil society representatives from more than 100 countries. This was the third global AI summit and marked a shift in focus from previous summits in the United Kingdom (UK) and South Korea, which had primarily discussed AI safety. The Paris summit, in contrast, focused on AI action, with discussions centering on AI in public service, AI’s impact on work, the role of AI in innovation and culture, the relationship between AI and trust, and how to best govern AI at the global level. Major speeches were given by government leaders from the U.S., China, France, India and the Czech Republic as well as by the United Nations (UN) Secretary General and the CEO of Google. Additional panelists included government representatives, such as the former Canadian prime minister, global tech industry representatives, including from OpenAI, Microsoft, and Mozilla, and academics, for instance from the Massachusetts Institute of Technology and University College London. 

Joint goals agreed upon – without the U.S. 

The joint statement from the AI Action Summit was signed by 60 nations and international organizations. The statement promotes a “sustainable and inclusive AI” and highlights the importance of international cooperation and regulation to address key issues, including AI’s impact on work, the environment, and society. It promotes AI trust and safety and highlights the importance of public-private dialogue in addressing AI’s energy needs. It also creates a number of initiatives, including a global “network of observatories to better anticipate AI implications” and a “public interest AI platform and incubator” to align national AI initiatives and identify AI governance inefficiencies. The United States did not sign the joint statement, saying it was opposed to AI safety measures, as well as to the statement’s general support for some form of global regulation. 

Global divergence regarding AI themes 

Global leadership was the most prominent theme in the summit’s major speeches, with most policymakers indicating their willingness and desire to lead in some aspect of AI. While some of these statements were provocative and departed in style from previous engagement in the AI governance field (EU, U.S.), others were more in line with traditional foreign policy behaviors (India, China, France, UN). Government actions at the summit reinforced these leadership claims. For instance, China hosted an event called “Progress in AI Technology and its Application” to highlight China’s AI successes and regulatory ambitions. 

The speeches also indicated general support for international collaboration. The EU was unique in prioritizing regional collaboration, while China and the UN Secretary General were the biggest proponents of global collaboration. Topical discussions focused on the need for global regulations, with some speakers calling for binding global rules (China, UN, India, Czech Republic), some rejecting such rules (U.S.), and some focusing on other regulatory aspects or settings (EU, France, Google). 

The theme of AI for global development was also prominent, primarily among speakers from the Global South (India, China) though also from the Czech Republic, the UN, and Google, which has much to gain from advancing AI projects and services in the developing world. Emphasis was placed on the education and training necessary to maximize gains from AI technologies (e.g., upskilling) and ensuring equal access to digital technologies and infrastructures. In contrast, panelists such as the CEOs of Mauritius’s Zindi Africa and Nigeria’s Amini pointed to structural barriers in AI datasets and training, as well as “diversifying the people who are building the AI solutions” as crucial challenges for improving AI’s developmental impact.

An additional theme dealt with the strong role for governments in AI development. This was most prominent in government discussions of their own investments in AI technologies and infrastructure (EU, U.S., France, India), but also appeared in Google’s appeal for governments to “take a thoughtful, strategic approach to AI to drive investment in infrastructure, people, and adoption.” Major speakers appeared to prioritize government strategic action in addressing AI challenges, limiting the private sector’s role to financial and technological assistance. Panelists disagreed, with the co-founder of Japan’s Sakana AI highlighting that private sector and open-source innovations could limit government control over AI developments and Mozilla’s CEO emphasizing the need to discuss the private sector as a positive force in maximizing AI opportunity.

Despite the prominence of AI’s need for energy at other recent global summits and in the AI Action Summit’s joint statement (see above), this topic was present, but not prominent, in major speeches at the AI Action Summit. France saw AI’s growing energy needs as a leadership opportunity, given French strengths in nuclear power generation. India, China, and the UN underlined the importance of sustainability and efficiency when addressing AI energy needs, while the U.S. promoted deregulation as a means of quickly developing necessary energy infrastructure. Panelists provided more detail. France’s TotalEnergies CEO discussed how AI can help improve energy efficiency and the CEO of AXA Group Operations argued energy optimization and innovation should be prioritized over managing potential problems arising from current energy projections.

Aggressive foreign debut by JD Vance

U.S. Vice-President JD Vance represented the U.S. at the AI Action Summit, delivering a forceful debut in his first foreign policy speech as Vice President. In his speech, Vance stated AI should be viewed as an opportunity rather than a threat. He opposed binding regulations at the national and global levels, which “could kill a transformative industry just as it’s taking off.” Vance described the U.S. as “the partner of choice for others” when using AI, but simultaneously underlined that U.S. efforts would focus on maintaining the U.S.’s position as “the gold standard worldwide” in AI technology. Along these lines, he rejected foreign AI partnerships, which could increase America’s vulnerability, and warned others against “chaining your nation to an authoritarian master that seeks to infiltrate, dig in, and seize your information infrastructure.” He also stated the U.S. would “center American workers in our AI policy” to increase American productivity and generate socioeconomic benefits.

This was the first of three controversial speeches that Vance gave in the European context over the next month. Like his subsequent speech at the Munich Security Conference and exchange with Ukrainian President Volodymyr Zelenskyy in the Oval Office, the speech at the AI Summit was met with mixed reviews. U.S. AI dominance has been questioned, given advances being made by China in other technological fields and the DeepSeek revelation, raising the idea that cutting-edge AI development is possible even without the deep pockets of Silicon Valley. Similarly, the tough stance Vance took towards international cooperation, while provoking some supportive action among allies such as the UK, raised the risk of American isolation in the AI realm, potentially creating new leadership opportunities for China. 

Yet others highlighted that Vance’s remarks were in line with the political priorities of the Trump administration to date and were intended to advance those initiatives abroad. Industry contributions implied Vance’s message - as well as the administration’s priorities - resonated at home as well, with Google’s CEO and Mozilla’s Executive Director both echoing Vance’s themes of opportunity, global competition, and innovation while depicting AI security and safety as technological rather than political challenges.

AI Action Summit Reference | Statements, speeches and panels

• U.S. Vice-President JD Vance: Full text

• (Joint) “Statement on Inclusive and Sustainable Artificial Intelligence for People and the Planet”: Full text

• Co-Host, French President Emmanuel Macron: Summary (no full-text available) | Video

• Co-Host, Indian Prime Minister Narendra Modi: Full text

• EU President Ursula von der Leyen: Full text 

• President of the Czech Republic Petr Pavel: Video

• Google CEO Sundar Pichai: Full text

• Chinese Vice Premier Zhang Guoqing: Summary (no full-text available)

• UN Secretary General Antonio Guterres: Full text

• Panelist remarks: Excerpts