The deputy Agustín Rossi registered on May 27, 2026 a request for information to the Executive Branch so that it details how the Social Digital Twin system will work, presented as an artificial intelligence tool applied to social policies.
The initiative focuses on a central contradiction. While the system is announced to manage social benefits, Rossi demands clarifications on whether it could end up automating sensitive decisions, such as benefit allocation, user profiling, or the termination of aid without sufficient human intervention.
Rossi asked for details on data, algorithms, and contracts of the Social Digital Twin
The draft resolution was sent to the committees on Constitutional Affairs and Communications and Informatics. In that text, the legislator requests information on the personal databases that will be integrated, the algorithms that will be used, and the type of artificial intelligence planned for the system.
He also requests to know if there will be external audits or any independent supervision mechanism. To this, he adds the IT security protocols, the allocated budget, and the contracts or agreements signed to launch the tool.
Furthermore, Rossi wants to establish whether the system will have the capacity to intervene in decisions that directly affect social benefit users. Among these assumptions, he lists the allocation of benefits, the generation of profiles or scores, and the eventual termination of benefits.
The project cites legal limits and questioned international precedents
In his proposal, the deputy warned that the initiative could involve massive and centralized processing of personal data of millions of Argentinians. Therefore, he argued that the State must apply a rigorous and transparent operating scheme before deploying a tool of such scope.
Rossi recalled that current legislation, including the Personal Data Protection Law, sets strict limits for the processing of sensitive information. In the same vein, he linked the debate to the need for democratic controls, independent audits, and human supervision over any automated system.
Outside of Argentina, the project mentions precedents that ended up under questioning. It cites the case of the Netherlands, where a social profiling system was declared illegal, and also experiences in the United Kingdom that received criticism for errors and biases in the allocation of benefits.
The text presented by Rossi maintains that Congress cannot remain on the sidelines of the development of technological tools of this type without adequate guarantees, and it was formally sent to the committees of Constitutional Affairs and Communications and Informatics.