Can artificial intelligence improve the administration of justice?

The opportunities, challenges and difficulties of Artificial Intelligence (AI) are inextricably linked to the asymmetrical development that is often inherent to countries in the Latin American and Caribbean region. It is very difficult to develop an inclusive and sustainable ecosystem, with automation and massive application of AI systems, while organizations are still working with processes and technologies from 70 years ago. Structural poverty, problems in accessing essential services and lack of infrastructure coexist with very embryonic digitalization, connectivity and interoperability ecosystems.

While new technologies are valuable tools in ending the “bureaucracy of paper,” and putting us on the path to a Smart State 4.0, many of the region's advances are hampered by the challenges facing the digitalization of processes in public organizations. Despite these challenges, Latin American justice has developed AI solutions, which seek to overcome these challenges and issues. 

To help understand the possible impact of AI adoption on justice application in Latin America, the Vice Presidency of Knowledge, State Digital Innovation Directorate, CAF – development bank of Latin America –, is leading a regional initiative to analyze the use of Artificial Intelligence as it is applied to the justice sector and evaluate its potential to reduce and manage three major characteristics of our societies: complexity, uncertainty and unpredictability.

Strategic use of data and artificial intelligence in the public sector

Data governance is the “oxygen of automation” and the very basis for the application of all AI systems. At the end of the day, the goal is to automatically process data to solve problems, assisting officials to complement and improve human competencies.

In the public sector, the potential of AI can provide high-impact solutions in the fields of health, education and justice, as well as in public employment, security and the management of citizen relations in general. It additionally allows greater accessibility to public services, while substantially reducing their costs by enabling the automation of public processes and procedures, offering citizens the opportunity to interact with the State in a more flexible, effective and personalized manner.

The introduction of AI based on proper governance data not only has the potential to optimize the swiftness of justice, but also allows collateral damage mitigation, while improving the efficiency of other powers and agencies linked to the judiciary, to the extent that it could allow the goal of a fairer and more equitable treatment of citizens to be achieved with regard to these issues.

Ethical challenges of applying artificial intelligence in the public sector

The use of AI in public sector decision-making processes requires in-depth reflection in terms of ethics and more broadly, governance. However, AI technologies can increase the potential for discrimination, negative biases, and may do disproportionate harm marginalized communities. For this reason, efforts must be made to ensure that the data sets used by AI system algorithms during the training, testing and operation phases do not contain unintentional biases, inaccuracies or errors, due to incompleteness or the models in use . To this end, it is essential that the data governance stage ensures compliance with quantitative and qualitative criteria, as well as data integrity.

The Organisation for Economic Co-operation and Development (OECD and the European Commission's high level AI experts recognize the essential requirements that AI systems must meet in order to be considered reliable: (i) human intervention and supervision; (ii) robustness and security; (iii) privacy and data governance/management; (iv) transparency and explainability; (v) diversity, non-discrimination and equity; (vi) social and environmental welfare; and (vii) responsibility and accountability.

AI experiences in Latin America

The research carried out places special emphasis on experiences in using these emerging technologies in justice throughout Latin America and the Caribbean. Due to asymmetries between growing demand and the people available to meet it, there is an increasing need for assistance with this type of technology in the region. It is to be additionally noted that the speed and adaptability of justice is one of the main requirements citizens have.

The study on the strategic use of data and artificial intelligence (AI) in the public sector to create social and economic value, soon to be published by CAF under the Vice Presidency of Knowledge, State Directorate of Digital Innovation (DIDE), identified a series of good practices and lessons learned that must be taken into account when designing, developing or implementing AI solutions in the public sector: 1) be clear that objectives must be oriented toward making technologies inclusive, 2) that prior to the design or development of AI systems, a diagnosis of the organization in question is to be performed, 3) that workers develop new skills and specific digital capabilities, while unlearning techniques, formats and approaches used under the previous paradigm, 4) that the system is designed with the user in mind, in an intuitive and friendly format, in a clear and understandable language available to all, 5) that human control is present throughout the system’s life cycle and that finally, procedures or formulas from other countries with different realities, cultures, economies and ecosystems, could only produce an effect contrary to what is desired.

In conclusion, the automation of tasks, documents and the simplification of public procedures in the justice sector could help legal actions be concluded in a shorter period of time, making justice more efficient and effective. This is how the democratization of AI benefits will contribute to a fairer State response to growing citizen demands.

On September 15, the ExperiencIA, data and artificial intelligence in the public sector forum will be held , where this and other examples will be discussed in depth, alongside executives responsible for this and related areas.