With 75% of Spanish companies prepared to deploy AI agents, organizations are no longer asking if they should deploy agents, but rather what workflows justify the cost of operating them and how to bring them to production. This is one of the main conclusions of “Agentic AI: The new human-machine alliance”, the latest report from MIOTI Tech & Business School, prepared from the responses of more than 70 Technology and Human Resources leaders from 25 sectors.
According to the study, which analyzes what companies expect from these systems, what risks they perceive and to what extent they are willing to delegate to them over the next two years, 90% of executives want agents capable of learning and remembering the work context, while 73% are still concerned about the operational hallucinations of the models, a growing concern over the risk that an artificial intelligence (AI) error ends up becoming a real action.
In terms of capabilities, beyond persistent memory, autonomy and control by agents, the study shows that companies are already beginning to demand additional and increasingly sophisticated functionalities. Multimodality, which allows working with different types of information, such as text, images, audio or video, is indicated by 7 out of 10 of the participants as one of the fundamental keys for the future, while 5 out of 10 highlight the ability to execute and orchestrate tasks.
The leap from Generative AI to Agentic AI greatly changes the dimension of existing risks. Until now, in a primarily conversational flow, the response generated by the AI normally reached a person, who could review it before using it or make a decision from it; However, when an agent has sufficient autonomy to act, the error could be transferred to a real process without prior review. Thus, concerns arise related to the control of information and decisions: 69% of leaders point out the uncertainty of where the data that AI is fed end up with and 59% point to the lack of existing traceability. To find a solution to hallucinations, 43% point to the government of knowledge as a necessary basis to avoid them.
“Until now we have worked with Artificial Intelligence that helped us generate, analyze or respond, but agents introduce an important difference: they can act,” says Fabiola Pérez, CEO and co-founder of MIOTI. “For companies this opens up many possibilities, but it also requires knowing what information an agent is working with, understanding why they have made a certain decision and clearly defining how far they can go without human intervention. Autonomy has to advance hand in hand with trust and control.”
Agent Managers: the new role of the human
The greater autonomy of AI agents does not necessarily mean that companies are considering removing people from processes. Far from this redistribution, what begins to change is the role that humans will have within them. As certain tasks can be delegated, professionals will gain weight in other essential tasks such as defining objectives, supervising work and providing criteria.
In fact, MIOTI points out in its report that 7 out of 10 leaders intend to keep the human in a role of supervision or control of the agent, not execution. Given this new role of human personnel, 37% identify a leading role as a strategic supervisor to evaluate the final result and 33% as an operational controller capable of authorizing each critical step. To achieve this, 47% focus on the reskilling and upskilling of professionals towards advanced prompting training and agent management. In this sense, the company predicts that humans will go from “People Managers to Agent Managers”, a new profile capable of coordinating the work of teams that no longer only manage people, but also agents.
New business: where to put your efforts
In this emerging panorama, the next step will be to identify which tasks these systems can take on from start to finish, which ones require validation points, and at what times human intervention is still essential. The transformation is already beginning to be reflected in companies’ priorities, and 75% of leaders point to the redesign of workflows as one of the main areas in which to invest. In practice, companies will have to decide what can be delegated, what information an agent needs to act correctly, and where human controls should be maintained.
Data and internal knowledge thus acquire a particularly relevant role: the greater the autonomy of these systems, the more important it is that they work on reliable, updated and correctly governed information. MIOTI precisely places data, architecture, governance and people among the necessary foundations to move from experimentation with agents to their real integration into the business.
The transformation is already beginning to be reflected in the companies’ priorities, and 75% of leaders point to the redesign of workflows
And not all areas will start from the same point. 39% of managers expect the first success stories to occur in operations and finance and 31% in marketing and sales, ahead of other areas such as human resources, legal, logistics or software development. Rather than extending agents equally throughout the organization, the report proposes a progressive evolution, starting with specific processes in which they can demonstrate their usefulness before escalating to other departments.
“We are entering a stage in which incorporating artificial intelligence no longer simply means giving a new tool to teams. Agents are going to be part of the processes themselves and that forces us to think again about how work is distributed, what we can delegate and where we need to maintain human judgment. The challenge for companies will be to combine technology, knowledge and talent so that this autonomy really translates into business improvement,” concludes Fabiola Pérez, CEO and co-founder of MIOTI.
