Artificial intelligence trains its own replacements using the voices of employees it is supposed to help

Artificial intelligence trains its own replacements using the voices of employees it is supposed to help

https://www.apd.cat/en/data-protection/artificial-intelligence-trains-its-own-replacements-using-the-voices-of-employees-it-is-supposed-to-help_12261_102.html

Publish Date: 2026-05-15 01:10:00

Source Domain: www.apd.cat

  • Dual Impact of AI Expansion: The growth of artificial intelligence in companies is reducing costs while simultaneously threatening jobs in administrative tasks, customer service, and routine management.

  • Reduction in Young Workers: Since 2024, a 16% reduction in young workers has been observed due to companies finding AI to be faster and cheaper than traditional methods.

  • Data Used to Train Bots: Expert María Luz Rodríguez highlights that workers’ voices, recorded during calls in call centers, are being used to train bots, thereby training technology that may replace human operators.

  • Ethical Use of Worker Data: Rodríguez stresses the importance of determining whose data is used to train AI algorithms, emphasizing that employment contracts often do not cover data capture, and calling for labor representatives to partake in technology governance.

  • Widespread Labor Market Impact: Rodríguez forecasts that AI will affect 70% of jobs, creating a disparity between high-value tech jobs and low-wage care jobs, suggesting a transformative shift in the labor market.

  • Human Skills as Protection: Rimbau argues that human transversal skills, like questioning and critical thinking, can serve as a safeguard against automation, highlighting the importance of these skills in a tech-driven workforce.

  • Regulatory Response: European regulations, such as the Digital Rights and Principles, aim to mitigate the adverse effects of AI, influencing global policies through the ‘Brussels effect’.

  • AI Agent Graveyards: Rimbau warns about the phenomenon of organizations creating AI agents that are neglected once the purpose changes, pointing to the need for sustainable technology development.