Approved Draft Bill for an Ethical, Inclusive and Beneficial Use of Artificial Intelligence

Approved Draft Bill for an Ethical, Inclusive and Beneficial Use of Artificial Intelligence

The Draft Bill for an Ethical, Inclusive and Beneficial Use of Artificial Intelligence is an initiative that puts Spain at the forefront of European regulation on emerging technologies.

Main Propositions of the Draft Bill

The Draft Bill includes guiding principles and specific obligations based on the AI Regulation, but which also add elements specific to the Spanish context, highlighting especially:

Mandatory ethical principles

The Draft gives vital importance to algorithmic transparency, explainability, proportionality of use, respect for fundamental rights and algorithmic non-discrimination. Any system that makes relevant automated decisions must incorporate human review mechanisms and guarantee traceability, so that at all times the tracing of information and the “humanization” of decisions can be traced.

Classification by risk levels

Following on from the IA Regulation, the Preliminary Draft classifies IA systems according to minimal, limited, high and prohibited risk. High-risk levels will have to undergo specific impact assessment controls on the fundamental rights of data subjects prior to implementation.

Promoting inclusive AI

The Draft Bill aims to promote AI systems based on social inclusion and the reduction of digital divides.

Impact on companies

All actors developing AI systems will have to implement systems and measures similar to those of the GDPR.
Entities that develop or implement high-risk AI will have to carry out the corresponding impact assessments and the aforementioned “humanization” of the decisions taken, as well as carry out audits of their algorithms so that the traceability of the data used to train the different AI systems can be verified.

Impact on citizens

Citizens will see their rights strengthened, especially with regard to being informed of the moment in which they act with AI systems and not with humans.

They will be able to request easily understandable explanations of the decisions taken, and they will be able to challenge those automated decisions that are considered to have significant effects, in a similar sense to that established for personal data protection regulations.

How your business can adapt to the new legal framework

Some keys for businesses to adapt to national AI regulations are:

• Conduct audits to detect which AI solutions are in use and be able to classify them according to the expected and tolerable level of risk.
• Conduct ethical impact assessments of all medium- or high-risk AI systems, including bias and discrimination risk analysis.
• Explainability of decisions should be facilitated, i.e., provide citizens and stakeholders in a simple and understandable way with the impact and rationale for decisions, and ensure that this is properly documented.
• Training and awareness of ethical use of AI within organizations should be ensured, and voluntary certifications should be sought to demonstrate a genuine interest in carrying out actions in the right way.

At Letslaw we can help you implement the legal requirements of your AI systems and ensure your company’s day-to-day compliance.