How iFood balances innovation and ethics in its AI models

In this article, Camila Nagano and Rafaela Câmara, experts on the topic at iFood, reveal the 3 pillars of responsible use of AI

iFood uses automated decisions and predictive models based on machine learning to manage everything from simple routine company activities to our complex logistics system. Today, the company has 124 artificial intelligence models.

Some of these models may use personal data that is protected by the General Data Protection Law (LGPD), and their use can significantly impact our ecosystem. The need then arose to implement a governance program for artificial intelligence (AI), with the aim of ensuring that models are designed and applied in an ethical and responsible manner.

For the program to be effective and, at the same time, allow technology to be explored in an agile way, we combine the actions of the professionals involved according to three pillars: culture, guidance and management.

The investment in culture institutional This occurs through training and internal alignment on the appropriate use of technology. Fostering a culture for the use of responsible AI allows employees themselves to question the impact that that model can cause and potential biases that may have in its construction. With this awareness, it is possible to anticipate potential risks and mitigate them.

Looking at the guidance of these professionals, it is necessary that they not only understand how the technology works but also teach them what privacy and ethics are. This is done through the training of models and algorithms based on ethical foundations that dialogue with legal privacy principles, such as: transparency, explainability, non-discrimination, security, robustness, responsibility and accountability. Using this approach, it is possible to provide AI with elements that enable fairer future decision-making.

Once the investment has been made in the culture and guidance of professionals who deal directly with technology, it is essential to manage these efforts to consolidate the AI model governance program. Yours management implies the adoption of effective measures that support the diligent use of AI, recording its activities, formalizing legal recommendations and monitoring its development and application.

It is a constant and lively work in the company's day-to-day life that must keep up with the rhythms of technological development, the market and regulation. The advancement of artificial intelligence, with emphasis on the growing use of generative AI in everyday life, makes it essential that ethics and privacy accompany innovation from the beginning. After all, we're just getting started!

Was this content useful to you?

Related posts