The Senate continues to advance in the incorporation of Artificial Intelligence (AI) into parliamentary activity and has opted to contract up to 65 licenses for “Gemini”, Google’s solution, for an estimated amount of 59,500 euros, according to the contracting document to which Europa Press has had access.
In February, the Upper House approved guidelines for the use of AI in its facilities, becoming the first Spanish parliament to establish specific regulation for this technology.
In this new step, the institution headed by Pedro Rollán promotes the practical deployment of Artificial Intelligence through a first package of 65 Gemini Enterprise Standard licenses, which also includes expert support services specialized in this tool.
The contract to have these licenses will have an initial term of one year, with the option to extend it annually.
The guidelines that mark the use of AI in the Upper House
Specifically, the Senate agreed in February to encourage the use of AI by senators in the performance of their parliamentary duties, always under ethical and legal principles, including human supervision and privacy protection.
These guidelines, applicable to both senators and all personnel working in the Upper House, aim to promote and facilitate the use of AI in the Senate, while at the same time establishing the ethical and legal criteria that must govern its correct use.
The Chamber has equipped itself with internal standards that regulate the purchase and deployment of AI systems that “maximize” respect for the ethical, legal, and organizational principles set forth in said guidelines.
To comply with this, Senate staff must undergo continuous training that allows them to identify the risks and limitations of the AI systems they use, as well as to adequately understand their capabilities and functionalities.
Risk assessment and information protection
In July, the Senate will host a meeting of presidents of regional parliaments focused on AI, where challenges such as the management of confidential information will be discussed. In that forum, the Upper House will share its experience to ensure that, under “no circumstances,” information protected by parliamentary secrecy is introduced into an AI system.
Among the approved guidelines to regulate the use of AI is that information subject to parliamentary confidentiality because it is not yet public, as well as that protected by intellectual or industrial property rights, will not be incorporated into public Artificial Intelligence systems. It may only be processed in AI systems authorized by the Senate that operate with non-public data when strictly necessary and, in any case, after prior consultation with the competent hierarchical superior.
