Reasons for requesting a pause on the AI Act from the perspective of a climate advocate.

      “There is no stopping the clock. There is no grace period. There is no pause.” This statement was made by Thomas Regnier, a spokesperson for the European Commission, in reaction to requests for a delay on the EU’s AI Act. These significant regulations regarding artificial intelligence will proceed based on the timeline outlined in the legislation. The European Commission will dismiss appeals from 46 CEOs across Europe, myself included, to postpone the legislation for two years.

      This may surprise some. My company, Kayrros, was the first globally to monitor and attribute methane emissions. We were referenced when John Kerry unveiled the Global Methane Pledge, and we contributed to the formulation of the EU’s methane law set for 2024. We have collaborated with numerous organizations to track greenhouse gas emissions and other climatic issues, such as deforestation and wildfires, as well as with many businesses eager to minimize their carbon footprints. Isn’t it true that unregulated AI could lead to a surge in energy consumption and the establishment of energy-intensive data centers?

      Here lies my concern. In its eagerness to enact the first significant AI regulation— which is driven by commendable intentions—the European Commission is losing sight of the bigger picture. Yes, we require safeguards. Yes, AI must be ethical and centered on human values. Yes, AI is an influential technology, and such powerful technologies can evolve in unforeseen ways. And yes, less regulation leads to increased AI implementation, which consequently results in higher energy consumption. Nevertheless, the AI Act remains a misstep.

      Firstly, AI has the potential to enhance industrial efficiency significantly, thereby reducing emissions for years to come. As noted by Nvidia’s Jensen Huang, once developed, AI models execute tasks far more effectively than conventional computing approaches. For instance, an AI-driven weather model can predict outcomes over 1,000 times more efficiently than traditional methods. Therefore, the energy used to develop AI is compensated—often with additional benefits— as the model is reused.

      Recent advancements in AI, particularly the emergence of more advanced large language models (LLMs) and remote-sensing foundation models (RSFMs), have transformed data analysis from a process taking days into one completed in mere hours. In my field, this translates to near-real-time analysis of satellite images, enabling us to forecast events like floods and wildfires, leading to significant cost savings and, in emergency service scenarios, potentially saving lives.

      However, as my colleagues at BNP Paribas, AXA, Siemens, and many other prominent companies have compellingly argued, the primary concern regarding the AI Act is that it may hinder our ability to compete globally in the most vital technological sector. It’s important to recognize that other global powers may not share our hesitations. They could lead the charge into the future, as they have in other essential fields like space, creating a widening gap that will make it increasingly difficult for us to catch up. This is not what we need: Europe is rich in world-class talent eager to establish the continent as an AI leader.

      There is also a lack of clarity present. Founders of startups, in particular, are understandably concerned about the regulation of general-purpose AI models. What happens if there’s a fragmented set of rules across different member states? Won’t larger, affluent US tech companies have an easier time navigating these complexities compared to smaller firms with limited resources? The fear of violating regulations—and incurring steep penalties—could discourage startups from innovating and deploying AI solutions. That’s the last thing we want. They are ideally positioned to innovate rapidly.

      The case for defense here is as compelling as the economic argument. The global power dynamics are shifting, and Europe is acutely aware of its need to bolster its capabilities, as demonstrated by the recent commitment to increase defense spending to 5% of GDP. Those in the defense industry emphasize the crucial role that data, software, and technological advancements now play in conflicts. AI is intertwined with this. The danger lies in jeopardizing our efforts to create a modern military force capable of maintaining peace in Europe and safeguarding European interests and values. AI is not merely one sector of the economy; it is the type of technology that will soon integrate into nearly every area.

      I do not oppose AI regulation in some capacity. However, I signed the open letter because I believe we are acting too swiftly, potentially endangering the soft power, economic security, technological sophistication, and military strength of our continent through our eagerness to regulate.