The indra subsidiary Minsait has developed a proposal for companies to prepare to adopt AI In its operations and processes complying with the EU’s law. It’s about AIwhich aims to guide companies to understand the level of ethical vulnerability of their AI systems, to know the legal obligations that correspond to them, and to evolve them to concrete practices within the organization without representing a reduction in their productivity.
The EU’s law was approved in March 2024, and entered into force last August. This regulation, applicable to both European companies and those operating in member countries, establishes a series of obligatory compliance requirements and gradual application.
Thus, according to the regulations, since last February, the AI systems used by some of the practices prohibited by the European regulations had to fall into disuse. In addition, companies have to have a literacy and awareness plan. As of August 2025, obligations for general purpose ia systems will be applied, such as LLM.
In August of next year it will conclude its implementation with the requirement of the requirements to some high -risk systems, that is, to those who may represent a threat to security or human rights.
The Minsait methodology to facilitate companies to adapt to the law consists of three phases. The first consists of the individualized analysis of the AI systems available to a company. For this, your data, techniques, logic, context and impact are studied. It also ensures that it includes all the information required by regulations.
Thanks to this analysis, the team of experts in the person responsible for Minsait can cross the information obtained on these assets with the risk levels contemplated in European legislation, in order to set categories based on the impact of the AI system from the point of view of ethics and responsibility.
The second phase leads Minsait to make available to the company an itinerary of compliance practices in the use of AI within the organization. Including routine privacy controls, data and transparency government. These itineraries are specified in particular actions that meet the specificities of each AI system, with the objective of facilitating the implementation of the practices required by the law of AI.
The third and final phase develops considering the previous work. It raises the company with a combination of technical components or tools that allow increasing traceability, relevance and ease of use for all AI managers within the organization.
On the other hand, as a result of the Minsait alliance with the Spanish Association of the Digital Economy (Adigital), AI Action offers companies the possibility of demonstrating compliance with the obligations established by the EU’s law to the Reduced Risk Systems and Risk Something with the Certificate of Algorithmic Transparency of Spain.
It is designed by Adigital and recognized by the OECD in its catalog of reliable tools of AI. Also, within the framework of the AI Action, Minsait and Adigital initiative they have also created a certificate of neutrality, which have recently launched.
Leticia Gómez Rivero, head of strategy and Government of AI in Minsaithe pointed out that «The great challenge of organizations is to translate the regulations into concrete practices, which requires the support of accelerators, with which we already have, and a cultural change. That is why the work we are developing from AI Acttion is so important and that has a focus on three elements: strategy, actions and tools that support good practices«.
