The Argentine Agency for Public Information Access (AAIP) has published a guide for public and private entities on Transparency and Personal Data Protection for Responsible Artificial Intelligence.
The guide begins by acknowledging that, in addition to the benefits we are discovering from the use of Artificial Intelligence, there are also risks and concerns that need to be considered. In this regard, it offers a series of recommendations associated with incorporating the principles of transparency and personal data protection in the development and use of Artificial Intelligence.
Among the main risks, the guide highlights the potential for reproducing and deepening discriminatory biases, violations of privacy, and the lack of transparency.
As proactive responsibility guidelines to prevent these risks, the guide emphasizes the importance of the following:
- Appointing a Data Protection Officer (DPO).
- Assessing the impact of the project on personal data protection from the outset.
- Considering privacy by design and by default.
- Ensuring that the data source used to train AI systems is lawful.
- Protecting personal data by applying techniques such as minimization, anonymization, encryption, access control, data retention, and destruction.
- Monitoring the quality of data and metadata.
- Ensuring traceability, auditability, and explainability of AI systems.
- Documenting processes, tests, and evaluations.
- Publishing a transparent privacy policy.
We hope that soon the relevant regulations will be enacted so that the good practices recommended by the guide, which are so important, become enforceable legal obligations!
