📋 Tender Overview
Public servants responsible for creating or hiring artificial intelligence (AI) solutions now have another ally to guarantee the positive effects of their work: the Ethical Impact Self-Assessment Framework, a test version of which was launched by the Ministry of Management and Innovation in Public Services (MGI) during Innovation Week 2025, which took place between September and October. It is a tool to assess the risks and ethical effects of an AI project and suggest changes before the adoption of the technological solution by the public administration.
This framework will help employees in the area and agencies to evaluate the consequences of adopting an AI solution in the public service, ensuring that the technology is applied responsibly and for the benefit of society, stated the Secretary of Digital Government, Rogerio Mascarenhas. This is important because it is one of the first documents from the Federal Executive providing guidance on this issue, an innovation that can even be used to regulate AI in Brazil, he added.
In practice, the framework will serve as a roadmap of questions that an AI project team needs to answer to, for example, ensure that the solution does not harm citizens fundamental rights. The process works in two main steps:
1. Risk Assessment: the team answers a quick questionnaire to identify the risk level of the AI project. Based on the responses, the system is classified as Low, Medium, High or Excessive risk (in this case, the suggestion is to interrupt the project for reassessment).
2. Main Questionnaire: Depending on the level of risk found in the previous step, the team is directed to a more detailed questionnaire.
At the end, the tool generates a report that offers practical recommendations to improve the identified points and suggests courses and training materials. The objective is to encourage teams to reflect on the impacts of their AI solutions, preferably before their adoption in the public service.
Currently, the framework is still in its beta version, but it can now be used on a test basis. For now, we are looking to mature the tool and, here at MGI, we will also apply it to some government AI solutions explained the digital government secretary. He also highlighted that this framework was built based on good international practices and considering the current discussions of PL 2,338/23.
PBIA and IND
The self-assessment framework for the ethical impact of artificial intelligence solutions is a project by the Digital Government Secretariat (SGD) of MGI that is part of a larger initiative: the Brazilian Artificial Intelligence Plan (PBIA). Launched in July 2024, this plan foresees an investment of more than 23 billion reais in four years.
The objective of the PBIA is to make Brazil a global model of efficiency and innovation in the use of AI in the public sector. To this end, solutions will be developed that significantly improve the supply and satisfaction of people with services, with an impact on development and social inclusion.
Axis three of the PBIA deals with the adoption of AI in public services and foresees an investment of around R1.76 billion. Projects linked to this axis, such as the recently launched framework, are also part of an initiative that seeks to increase security, interoperability and improve the use of federal government data, the National Data Infrastructure (IND).
Also known as Brazils database, the IND is a set of norms, policies, architectures, standards, technological tools and information assets that aims to promote the strategic use of data held by bodies and entities of the federal executive to expand citizenship with transparency and security and facilitate access to social benefits and public services, focusing on the motto a government for each person.
