The President of the United States, Donald Trump, has reopened the international debate on artificial intelligence by suggesting that the Government should have a "kill switch" capable of deactivating advanced systems in case of risk.
Although no concrete legislative proposal has been formalized, the idea is framed within the growing fear of the so-called existential risks of AI, a concern that has been gaining weight in recent years both in the political and scientific spheres.
A proposal in a context of political blockade
Trump's reflection comes at a time when his technological agenda is facing difficulties in Congress. The White House has tried to promote a federal framework for artificial intelligence, but political divisions have slowed its progress.
In that scenario, the proposal of a total control mechanism over AI introduces a new element into the debate: to what extent the State should have the capacity to intervene directly over increasingly complex technological systems.
The “shutdown button”: between security and uncertainty
The concept of “shutdown switch” —known in the technological field as kill switch— is not new. It is a mechanism designed to stop a system immediately if its behavior becomes dangerous.
However, its application in artificial intelligence raises technical and ethical doubts. Experts warn that the most advanced systems could be difficult to control or even resist a disconnection if it conflicts with their programmed objectives.
Furthermore, many of these systems operate in a distributed manner, which complicates the existence of a single effective control point.
Trump's position reinforces a political line focused on security and state control of strategic technologies, in a context of global competition with powers like China and of growing concern over the impact of IA on the economy, disinformation or defense.