When AI goes wrong, who validates?New profession emerges to oversee intelligent agents

The development of autonomous agents continues to redefine the global labor market as one of the most significant advances of the Artificial Intelligence era. The adoption of AI Agents has driven the demand for new functions aimed at the supervision, validation and quality assurance of automated tasks. According to a recent report by the World Economic Forum (WEF), by 2030, about 22% of current jobs will be transformed. The projection points to the creation of approximately 170 million new jobs and the extinction of 92 million jobs, resulting in a positive balance of 78 million jobs. 

 
For Marcos Oliveira Pinto, Global Software Engineer Manager at Jitterbit, the profession of validator of Artificial Intelligence agents emerges as a strategic function in the future of intelligent automation in business. “We are still at a time when it is not possible to blindly trust AI. It may seem obvious, but it is essential to reinforce the need for human presence in processes and in the creation of new functions in the market. The AI validator can easily be one of them, ensuring efficiency with real results”, he explains.   
 

The attributions of this new function go beyond risk mitigation. The role of the validator involves ensuring a safe, ethical and reliable management of operations in companies that adopt intelligent agents as part of their processes. “The main concern is in the way these systems are built and the possibility of AI generating inaccurate or incorrect answers. The objective of the professional will be to review, validate and confirm the actions of the agents, ensuring the absence of hallucinations and the expected return of the tasks performed”, says Marcos. 

According to the executive, the integration between humans and intelligent agents can follow two distinct models. “In the active model, the human participates directly in the execution. The agent conducts the task to a certain point, such as in the service to a user, and from there the professional evaluates and decides the next steps. Already in the passive model, the human receives notifications about the progress of the operation. In both cases, the human presence is essential for approvals, workflows, strategic decisions and critical judgments”, he details. 

The role of the human employee becomes to ensure the success, safety and reliability of automated stages of the journey, especially in activities that require greater complexity, persuasion or technical knowledge, still heavily dependent on human perception or hybrid approaches. 

The World Economic Forum survey also highlights that the skills most valued by employers remain the so-called core competencies. Analytical thinking leads the demand, followed by human attributes such as resilience, agility, leadership and social influence. By 2030, the fastest-growing competencies expected include technical areas such as AI, big data, networking, cybersecurity and technological literacy, as well as creativity, flexibility, curiosity and continuous learning. 

“O validator of AI agents synthesizes the essence of the new era of work.This is a professional who combines technical knowledge in artificial intelligence and new technologies with the human judgment necessary to deal with uncertainty and ensure confidence in autonomous systems. This function reflects the trend that agents will need to be trained and supervised by trained humans, able to recognize critical situations and apply critical sense.In the end, our differential remains the attentive look and continuous development, regardless of the profession”, concludes Marcos.