“If organizations do not address data bias, they can suffer irreversible long-term legal, security and reputational consequences,” warns Larez, who recommends that companies pay more attention to the interest in AI in all its spheres.
Larez comments that although it is a new subject and understanding the data is complex, he accepts a greater tendency to become aware of it. Therefore, a fundamental ethical framework must be established and, as adoption increases, bet on more regulations.
How to deal with AI biases
Given this panorama, the specialist provides some recommendations for companies that are in the process of minimizing biases. The first is to become aware that the problem exists and from there establish the best sources of information and that the samples are sufficiently representative to avoid discrimination.
Another point that stands out in this regard is that it must be understood that the biases are dynamic, that is, they are not resolved when they occur and the gaps of the present are not the same as those of the future, so a process of constant monitoring of the data that is integrated into the tool.
On the other hand, the dilemma of what the adoption of AI could mean in the labor market arises and there is no doubt that they will do routine jobs. One of the things that it will promote is that professionals should dedicate themselves to more strategic things in the business.
