This Sunday has entered into force the first phase of the European Artificial Intelligence Regulation (AI), approved in August 2024 that prohibits companies from using such technology to recognize emotions in jobs, which will be fined with up to 35 million euros or 7% of your annual business figure.
Thus, the use of IA devices or applications that use subliminal techniques aimed at influencing the behavior of people, who take advantage of vulnerabilities, which serve to categorize individuals or to identify them through biometry, among others, will not be allowed. These prohibitions concern all companies, although some more than others.
You can’t know if a worker “is motivated”
These prohibitions concern all companies, although some more than others. In this first phase, those of the Security and Analytical, Human Resources or Sports Entities sector will be the most affected, the president of the Commission on AI of the Employers of the Ametic technology industry, Enrique Serrano, points to Efe. Thus, for example, AI cannot be used to know if a worker “is motivated at work,” he says.
Take the opportunity to ask institutions support so that companies can comply with the regulations: there have been aid to incorporate AI systems, but not for the adoption of measures for this law, he adds.
In addition, the suppliers and responsible for deployment of AI systems must take measures to ensure that their staff has a sufficient level of literacy in this area, as recorded in the regulation.
The unions, vigilantes
From UGT, José Varela, responsible for AI and digitalization, has warned that unions will be vigilant for companies to comply with these regulations.
At the moment, today Chapter I of the Law (general provisions) and Chapter II (prohibited practices) enters into force. The rest of the precepts will be applied in several phases until reaching February 2, 2027, when it will already be in force in its entirety. In just two years, any type of job use must be “explainable, transparent, human controlled and must overcome risk and quality audits. 35 million or, if the offender is a company, up to 7% of its total world business volume corresponding to the previous financial year, if this amount were superior. It will be the Spanish Supervision Agency of the AI (AESIA) in charge of raising the fines.
There are still no concrete figures for the impact of this law on the company, since most of them refuse to provide specific data on the type of AI tool that they have running in their work centers, says José Varela.
In addition, for the moment, the use of AI is not widespread, taking into account that the number of SMEs that use the IA does not reach 3% and 12% in the case of large companies. Many of these prohibitions of the law are complemented by the recent opinions of the Spanish Agency for Data Protection (AEPD), which, for example, prohibits using the fingerprint or any other biometric data as a method to register the day.
#prohibitions #European #Law #Artificial #Intelligence #fines #million #enter #force
