Prosegur: Shaping an Ethical Future For AI in Security

This system, a pioneer in the Spanish corporate landscape, includes the company's first Responsible Artificial Intelligence Policy, a series of guidelines based on respect for human autonomy and principles such as fairness, transparency and prevention of harm. Although in Spain, for now, only a small percentage of companies have protocols of this nature, in the near future, responsible AI will become the norm, driving ethical and transparent solutions that generate tangible positive impacts for society as a whole.

With the intention of highlighting the company's ongoing commitment to cutting-edge technological innovation, we spoke with Diolimar García, Daniel Bastida and Carolina Cortés: Prosegur’s Legal Director, Data Protection Officer, and Innovation Manager respectively, and all members of Prosegur's Responsible AI Committee. These experts uphold the company's firm commitment to fuse innovation, technology and ethical values in the creation of advanced security solutions.


Outlining Ethical and Technical Criteria


As part of the Compliance Unit, the Responsible AI Committee is responsible for making sure that all Prosegur AI solutions follow the guidelines set out in the policy throughout their life cycle. These projects must respect ethical and moral values, as well as being sufficiently robust, both from a technical and social point of view.

Diolimar García explains that the committee oversees compliance with AI requirements and the defense of  fundamental rights and freedoms. It also keeps a record of all the solutions implemented in Prosegur and advises the different areas on all legal and ethical issues affecting AI projects.

"With this policy of responsible use of AI we want to join other large companies that place technology at the center of their operations, and we intend to do so in an ethical, lawful and robust manner, as indicated by our guidelines," says García.


Unbiased innovation


Prosegur’s policy aims to provide an adequate response to one of the major concerns raised by the widespread use of AI: how to avoid unwanted biases when developing algorithms. As Carolina Cortés explains: "It is very important to follow a process aimed at guaranteeing the quality of the data, both in its collection and in its treatment, since biases can spread from the data to the decision-making process of the models."  

The committee has a specialized methodology for assessing compliance with the requirements set out in the policy. This not only ensures compliance, but also effectively contributes to minimizing the risk of bias in the AI development process and in the algorithms used.


Human supervision and technical robustness


Because Prosegur operates in the private security and technology services sector, which is a highly regulated and ethically demanding field that requires rigorous internal codes of conduct, it has drafted its policies  based on the ethical principles set out by the European Commission, highlighting respect for human autonomy.

As Diolimar García notes, "a sufficiently robust Artificial Intelligence must guarantee a minimum of human intervention and supervision in the development and use of the same." This way correct control of the systems can be established to prevent them from making decisions in a completely autonomous manner. 

"Although the application of AI brings great benefits, we must also prevent its risks. These systems have to be designed in a way that complements and enhances people's skills, and the only way to ensure their safety is through human supervision and control over the work processes of these systems," he explains. 

In addition, this regulation also ensures the technical soundness of the systems to ensure proper management of data privacy, preservation of diversity, non-discrimination and fairness. It therefore introduces transparency or accountability criteria for AI systems and their results both before and after implementation. 


Responsible AI in action: success stories


By aligning themselves with the most rigorous European compliance standards, all Prosegur's AI-enhanced products and services are positioned as ethical benchmarks in terms of regulatory compliance.

Last year, after the launch of its Responsible AI Model, Prosegur won Spanish business daily Expansión’s Best Ethical Initiative in Compliance award.  The model improves effectiveness and efficiency in the management of projects that incorporate AI technology, marking a turning point in the application of ethics in AI. 

Among the solutions proposed by Prosegur using AI technology that have successfully passed compliance with this Model of Responsible AI are a dog-robot able to detect risk events and send alerts to a control center, another robot capable of detecting people, obstacles or intrusions that can issue alerts for smoke, fire or abandoned suitcases. Other solutions include the G-Watch system designed to verify the correct operation of video surveillance cameras, and the G-Home Intelligent Recognition.

By successfully meeting the requirements of the AI Management Model and rigorously complying with the principles and requirements established in its Responsible AI Policy, Prosegur projects using AI technology, whether developed in-house or by third parties,  guarantee compliance with ethics and respect for the rights and freedoms of individuals, creating a benchmark for responsibility in the application of this technology.

In conclusion, the growth of AI in the business sphere has opened up a range of ethical dilemmas and challenges that require forceful and transparent responses. With these initiatives, Prosegur has taken another step forward in its purpose of caring for people and companies, demonstrating that technological innovation and ethical values can coexist to make a positive and sustainable impact on society.