Artificial intelligence has become one of the main arenas of global economic and geopolitical competition. In this contest, rapid development is rewarded, the ability to operate at scale is viewed as an advantage, and market leadership is a central objective. According to an article published by Project Syndicate, the risk is that technology’s ultimate purpose may become secondary.

The starting point is no longer whether AI will change the world. According to the article’s author, that transformation has already taken place. The question has therefore shifted from technical feasibility to strategic direction: society must decide who should benefit from this new capability and what criteria should be used to assess its consequences.

Against this backdrop, the concept of “Compassionate AI” was presented at the AI for Good Global Summit held in Geneva earlier this month.

The idea stems from a call to equip AI development with a moral compass. The source does not describe operational tools, binding rules or an established institutional model. It does, however, identify a priority: putting the people who should benefit from innovation back at the center and considering the human consequences of decisions made when designing and deploying these systems.

The proposal comes in an environment dominated by tangible incentives. Speed allows developers to move ahead of competitors, scale strengthens distribution capacity, and market control consolidates economic power. These forces drive the global race but, according to the argument presented, can divert attention from the fundamental question of what artificial intelligence is meant to do.

The appeal to compassion is not framed as criticism of technological transformation itself. AI is described as a reality already capable of reshaping the world, rather than a future possibility to be accepted or rejected. The central issue is how this transformation is governed and whether it is also assessed according to the people it serves.

The economic and geopolitical dimensions are closely intertwined. The importance assigned to scale and commercial dominance shows that the contest is not limited to research or system efficiency. It also concerns the distribution of the power generated by artificial intelligence and, consequently, the priorities guiding its use.

The audience at the Geneva summit gave the proposal a multilateral setting. The presence of delegates from 170 countries indicates that the debate extends beyond individual companies or national markets. At the same time, the available information does not specify whether the concept received formal endorsements or whether organizations, programs or commitments are planned to put it into practice.

It also remains unclear how a moral compass could coexist with the incentives created by global competition. The source identifies the tension between market dominance and serving people but provides no measurable criteria for determining when a system can be considered compassionate. It sets out no standards, oversight procedures or specific responsibilities for governments and businesses.

The next step will therefore be to determine whether the proposal remains a broad principle or enters the AI governance debate through more clearly defined instruments. For now, the message from Geneva establishes the basis of the discussion: after the race to build faster and more powerful systems, the question of which interests should guide their development and use remains unresolved.