How Will Agentic AI Impact the Job Market?

Following the boost that artificial intelligence (AI) has received from the popularisation of generative AI, thanks to ChatGPT and other similar platforms, this technology is now embarking on a new phase.

We are referring to the development of agentic AI or autonomous agents, which represents ‘the next evolution of automation,’ as Salesforce stated in this report a few weeks ago.

This technology inaugurates advances such as autonomous management or minimal human supervision of complex customer queries, automated and personalised product recommendations, real-time assistance, hyper-personalised digital marketing actions, intelligent complaint resolution systems, supply chain optimisation, improvements in health case management, internal employee support solutions, etc.

Impact on the economy…

This is a most welcome development for businesses, as they will be able to leverage agentic AI to improve their efficiency and productivity. As a result, the economy as a whole will benefit.

“Agentic AI has extraordinary potential to transform business operations, primarily through task automation, agile decision-making, and large-scale personalisation. At IBM, we estimate substantial improvements in operational metrics. For example, we project a 51% reduction in the order-to-cash cycle in commercial processes, as well as a 35% increase in employee productivity in areas such as human resources by 2027. These advances, combined with 24/7 touchless operations, have a multiplier effect on competitiveness and organisational efficiency,’ says Luis Sanz, Senior Partner & VP of IBM Consulting.

Likewise, Luis Miguel Garay Gallastegui, professor of the Master’s Degree in Artificial Intelligence at the International University of La Rioja (UNIR), points out that various studies agree that agentic AI will be one of the main drivers of global economic growth in the next decade.

‘According to PwC, its adoption could raise global GDP by up to 15% by 2035. And McKinsey estimates that it could generate between $6.1 and $7.9 trillion annually, which would add up to 3.4 percentage points to global productivity growth. Although the figures vary depending on the source, the consensus is clear: this technology has the potential to profoundly transform production models, massively increasing efficiency in all economic sectors,’ he explains.

Similarly, Gonzalo Vázquez, Chief Technology Officer (CTO) for Europe at Globant, states that ’agent AI has a huge potential impact on the global economy.’ ‘It is estimated that it could contribute between $2.6 and $4.4 trillion annually to global GDP, mainly by boosting productivity. By allowing intelligent agents to take on repetitive, analytical or even creative tasks, we free up time and resources for higher-value activities, which could improve efficiency in key sectors such as healthcare, finance, logistics and retail by up to 30-40%,’ he points out.

In line with this data, Teresa Castaño, strategic business advisor and HR expert at SAP SuccessFactor, states that the use of AI agents in areas such as customer experience, supply chain management, expenses, finance or human capital management leads to productivity increases of up to 30%, as well as increasing efficiency and promoting constant innovation.

In fact, beyond savings or optimisation, Vázquez emphasises that ‘agent AI opens up new opportunities to create completely different business models, where personalisation, real-time adaptability and the autonomy of solutions change the rules of the game’. ‘At Globant, we are convinced that this technology is a strategic enabler for companies that want to lead the next wave of digital transformation,’ he says.

In line with this, Castaño notes that ‘companies such as PwC predict that agentic AI could contribute $200 billion annually to the global economy by 2030, driven by the reinvention of business models.’

And it seems that companies are eager to rise to the challenge. According to the latest Microsoft Workplace Trends Index (WTI), 89% of Spanish executives plan to incorporate AI agents in the next 12 to 18 months. This figure places the country 12 points above the European average (77%) and 7 points above the global average, highlighting Spain as one of the most dynamic markets in the evolution of the labour model through AI,’ explains Antonio Cruz, director of Modern Workplace at Microsoft Spain.

…Also in the labour market

All these figures are very appealing. However, the adoption of agentic AI raises a big question mark regarding its impact on the labour market, as the deployment of this technology could lead to the disappearance or transformation of many jobs.

“According to market data, agentic AI is expected to grow at a compound annual growth rate (CAGR) of 40.3% between now and 2034. This implies a profound transformation of the labour market, where some professions will be forced to evolve to learn how to work with AI due to automation and improved efficiency, while new professions will be created in areas related to AI and technology,’ says the SAP manager.

It is therefore clear that many jobs will be destroyed, either through direct replacement or through efficiency improvements that result in job losses. Although no one dares to give figures. But new jobs will also be created, so it is worth asking whether this balance will be positive or negative.

‘Initially, it will be a distinctly negative revolution in terms of the number of jobs. But new job opportunities related to agentic AI will emerge. What will make the difference in terms of the impact on employment is whether companies move to create, develop or customise their own solutions or opt for SaaS-type solutions, developed and managed by others,’ says Rodrigo Méndez, CEO of Tecnoempleo.

Similarly, Patricia Zamacola, Director of Technology at LHH, an Adecco Group company, believes that ‘work will be transformed, departments will focus on new tasks and talent will focus on the more business-oriented side’. She therefore believes that AI will not destroy jobs, but will transform existing positions.

Castaño is also optimistic and believes that the result of this ‘new wave of transformation’ will translate into ‘the creation of jobs in emerging sectors and the transformation of existing roles that can offset losses, resulting in an overall positive balance.’

Similarly, the VP of IBM Consulting admits that the impact of agentic AI on employment will be profound, but believes it does not have to be negative. ‘The key is transformation, not disappearance. At IBM, we see that many of these functions will evolve to integrate with agent supervision, intelligent workflow design and strategic data interpretation, enabling professionals to elevate their work to higher value-added tasks,’ he notes.

For this reason, Castaño emphasises the importance of preparing for this transition. ‘To ensure this positive balance, it is important to opt for proactive government strategies to transform the composition of the labour market with reskilling and upskilling programmes, programmes to increase the population’s AI skills, as well as programmes to help workers affected by professions that will become obsolete,’ he points out.

‘The key is to prepare ourselves as a society. If we invest in education, in fostering digital talent and in promoting human-machine collaboration, agentic AI will be a lever for generating many more opportunities than losses,’ adds Vázquez.

Experience shows that the balance can be positive if we invest in training and adapting the workforce. ‘A study by the IBM Institute for Business Value points to a projected 2.5% increase in total workforce by 2026 in companies that are already advanced in their adoption of AI, indicating that, if well managed, this technological transition can coexist with employment growth,’ explains Sanz.

Which jobs will be most affected?

The emergence of new technologies and the uncertainty this brings is nothing new. It has been happening since the Industrial Revolution. And that experience allows us to know that the arrival of agentic AI, like any other advance, does not impact the entire labour market in the same way, but rather that some tasks will be more affected than others.

‘The most vulnerable jobs are those that are easily automated and add little value. In contrast, all those strategic, business-focused jobs that require genuine conversation will be the least affected by AI,’ says Zamacola.

The head of Globant points out that ‘the most exposed tasks are those that are routine, predictable and based on clear rules, such as data processing, structured information analysis, basic writing, customer service or even some administrative functions, where agentic AI can take on much of the work, freeing up time for higher-value tasks.’

For example, Garay mentions tasks such as reviewing invoices, sorting emails, transcribing minutes or managing schedules. ‘These can already be handled by intelligent agents,’ he says. He also points out that functions such as preparing basic reports or responding to frequently asked questions in customer service centres are being automated.

Similarly, the VP of IBM Consulting believes that customer service will be greatly impacted, since ‘AI can already handle more than half of calls that involve actions such as changing an address or recovering credentials.’ ‘It is expected that more than 70% of basic interactions will be handled by virtual agents in the next two years,’ he says.

In sales and marketing, Salesforce points out that ‘AI agents can respond to common queries, qualify sales opportunities and optimise marketing campaigns, reducing the need for human intervention in these tasks’.

The company also highlights the capabilities of agentic AI in the field of human resources. ‘AI agents can manage holiday requests, personal data updates, expense reimbursements and other frequent employee queries. These tasks, which previously required human intervention, can now be efficiently handled by AI agents integrated into platforms such as Slack or employee portals,’ it specifies.

Likewise, the SAP spokesperson indicates that agentic AI will impact ‘tasks susceptible to greater automation, such as record management, report generation, or data entry, reducing the need for traditional administrative roles.’ The same will happen in tasks such as inspection and quality control, especially in the manufacturing sector.

On the other hand, Vázquez notes that ‘the least exposed roles are those that require creativity, critical thinking, empathy, leadership and decision-making in ambiguous environments, such as designers, strategists, advanced technical profiles, as well as health and education professionals, etc.’

‘This is the case for customer and supplier relationship management, business strategy formulation, organisational design, ethical supervision of AI agents, and decision-making in uncertain contexts. In these cases, agents act as assistants or collaborators, not as substitutes,’ Sanz points out.

In which sectors will it have the greatest impact?

Agentic AI will reach all sectors, but its impact will vary substantially from one to another. ‘Its impact will be particularly strong in sectors that are intensive in data, complex processes and customer relations. For example, in financial services, we will see agents managing customer interactions, risk analysis and product customisation. In retail, they will revolutionise customer service, inventory management and demand forecasting. In healthcare, agents can optimise data management, support diagnostics and facilitate patient follow-up. Other sectors, such as logistics, education, human resources and legal services, will also see profound changes with the automation of repetitive tasks, allowing these professionals to focus on activities of greater strategic value,’ says the Globant representative.

Sanz also highlights the foreseeable impact on financial services, retail, manufacturing, telecommunications and healthcare, where ‘practical applications and tangible improvements are already being seen thanks to automation and real-time AI-based decision-making.’

“In finance, for example, improvements of more than 30% are expected in KPIs such as accounts payable cycle and forecast accuracy. In human resources, recruitment and employee self-service processes are being redesigned around virtual agents. In customer service and sales, multilingual assistants and dynamic personalisation are significantly improving the user experience in both B2C and B2B environments,’ he explains.

Cruz notes that ‘companies in industry, healthcare, financial services, law firms and many others are incorporating these agents.’ ‘In industry, they help optimise production processes and predictive maintenance. In healthcare, they facilitate the review and organisation of clinical reports. And these agents allow financial and legal companies not only to save time and costs, but also to make more informed and agile decisions,’ he specifies.

For his part, Castaño emphasises its foreseeable impact on manufacturing and production, customer service, supply chain management, finance and accounting, human resources, marketing and customer experience, technology and software development, and sustainability and environmental management.

New opportunities for agentic AI

The development of agentic AI will also give rise to new responsibilities and professional profiles. ‘The implementation of agentic AI is driving technical profiles such as AI engineers and architects, data scientists and cybersecurity experts, who are responsible for designing, training and protecting these systems. There is also demand for programmers specialising in integration, who connect agents with internal company processes, and cloud architects, who guarantee scalable and secure environments. Added to this are new roles, such as algorithm auditors and technology ethics consultants, who ensure the effective, transparent and responsible use of AI,’ explains the UNIR professor.

Méndez also highlights the emergence of profiles such as ‘data managers, who monitor the ethical and legal use of the data fed into these models; and user experience specialists, who ensure that users served by these AI agents have a satisfactory experience.’

Salesforce also mentions AI agent specialists, conversational experience designers, AI system integrators, data and AI ethics analysts, AI auditors, AI model trainers and AI user experience designers.

The Microsoft spokesperson notes that one of the profiles that will emerge with agentic AI is that of the agent manager, ‘focused on building, training, supervising and optimising AI agents’. ‘This figure directs interaction with these systems, guides them according to business objectives and translates their capabilities into concrete benefits. The need for this profile responds to a specific reality: AI is no longer a static tool, but a dynamic collaborator that needs guidance, objectives, adjustments and evaluation. They will be the ones who direct that interaction, ensure its alignment with the business and turn the technology into a real advantage,’ he explains.

In a similar line, Sanz refers to the agent orchestrator, ‘responsible for coordinating how multiple autonomous agents interact within a workflow’; specialists in AI ethics or operational impact analysts.

For his part, Globant’s CTO for Europe states that ‘we will see the creation of new roles that we cannot even imagine today,’ such as agent designers, experts in optimisation and orchestration of agent systems, experts in ethics and algorithm governance, or AI agent trainers, responsible for teaching agents how to interact, what tone to use, or how to adapt to different scenarios.

On 19 June, we will be holding our Silicon AI Leaders Day event, where experts from the technology sector will explore the current state of Artificial Intelligence and its adoption at the corporate level. If you would like to attend, simply register here. We look forward to seeing you there!