-
Óscar Méndez Soto on Capital Radio: governance, traceability and the future of AI
In his interview on Capital Radio's After Work program, Óscar Méndez Soto, president of GenAiA, analyzes the main challenges posed by the evolution of artificial intelligence and advocates for a governance capable of combining innovation, security, and responsibility.
-
The five mistakes companies make when implementing AI agents
The success of AI agents depends much less on the model than we usually think and much more on the enterprise architecture: data, security, governance, oversight, and process design.
-
Óscar Méndez Soto in Smart Factory Magazine: "Industry 5.0: Productive muscle or the best-regulated museum in the world?"«
Óscar Méndez Soto, president of GenAiA, analyzes in Smart Factory Magazine the challenge for Europe to turn artificial intelligence into true industrial capability.
-
Óscar Méndez Soto on Nació: "That the Manhattan Project of AI is in private hands is a path to disaster"«
Óscar Méndez Soto, president of GenAiA, analyzes in Nació the challenges of accelerating artificial intelligence and the need to strengthen European governance and technological sovereignty.
-
Óscar Méndez Soto on Hora 25: "We haven't reached the point of no return yet."«
The president of GenAiA analyzes on Cadena SER why the advancement of artificial intelligence demands an acceleration of security, governance, and control mechanisms. Is our capacity to govern artificial intelligence advancing at the same pace as the technology itself? Following recent warnings about the risks associated with increasingly sophisticated systems…
-
Governing uncertainty: when should AI abstain
There are labyrinths that are not traversed by walking, but by remaining silent. Artificial intelligence has become one of them: a mirror that multiplies possible paths and, at the same time, conceals the point at which the reflection ceases to be legitimate. Governing an AI is not simply about tallying successes. It is, above all, about establishing…
-
AI requires purpose
That a system functions and meets applicable requirements does not answer the fundamental question: why should AI be involved in a particular decision? The analysis posits purpose, oversight, and accountability as prerequisites for automation.
-
GenAiA warns that "the biggest conceptual error" is allowing the use of the public cloud in government.
In an interview with Europa Press, Óscar Méndez Soto, president of GenAiA, analyzes the challenge of technological autonomy for Spain and Europe in the face of an artificial intelligence increasingly linked to the control of critical infrastructures, data and computing capacity.
-
Evaluating generative AI requires looking beyond the model
Evaluating generative AI requires considering more than just model performance. Language, context, instructions, system usage, and behavior are all part of an evaluation that must necessarily be multidimensional.
-
AI will not be reliable if it is not representative
Representativeness cannot be incorporated once a system is finished. It begins by defining what problem will be solved, what data will be used, and who participates in the decisions that will determine its operation.