Geyto Research · Juan Ferreiro · Market note · · 2 min read

Using AI is not adopting it. What the data from Argentina shows

People already use AI at work; organizations have not adopted it yet. What recent surveys on Argentina show and what it means for companies.

In brief: in Argentina, AI has already entered the workplace, but it came in through people, not through organizations. Recent surveys show widespread individual use and organizational adoption that is still at an early stage: in most cases there is no strategy, no training and no measurement. The distance between using AI and adopting it is the same one that separates an installed tool from work that has changed.

People already use it

According to Microsoft, in the first quarter of 2026, 17.8% of the world's working-age population used a generative AI tool, and Argentina sits above that average. The Monitor Nacional de Inteligencia Artificial, presented in February 2026, found that six in ten people use AI in their personal lives.

At work, the CEPE Di Tella and Fundar survey of individuals (August–September 2025) found that 27.5% of employed people use AI for work tasks, and 89.1% of them feel it makes their work easier.

Organizations, not yet

On the organizational side, the picture is different. Among those who use AI at work:

  • 16% say their organization has an official AI strategy;
  • 6.2% say their company pays for the tool they use;
  • 13.7% received some training from their organization, and 85.6% learned on their own.

The authors describe it as "bottom-up" adoption. In the Monitor Nacional, 43% of workers say their company uses AI, but only 6% say it has been implemented broadly.

SMEs, at an experimental stage

The nadIA survey of 402 Argentine SMEs (November–December 2025) shows the same pattern from the company side. 41.6% use at least one AI technology, and more than 85% of them started in 2024 or 2025. But among those that use it:

  • 44.4% say individual areas or employees adopt it on their own, without central coordination;
  • 72.5% have no policy or usage guidelines;
  • 57.4% do not formally measure the impact.

The barriers companies themselves report are not, first of all, about cost. They are about capacity: a lack of experience or knowledge (46.2%) and of qualified staff (42.2%), ahead of high costs (29.2%).

What this says about adoption

The data describes a familiar situation in technology adoption, with one difference: this time the tool arrived before the plan. In many organizations AI is already in use, but in an individual, uneven and barely visible way. That has an upside, because there is genuine demand and curiosity, and a risk: the work does not change as an organization, and what each person learns is not shared or sustained.

Adopting AI is not enabling access. It is deciding which tasks change, who reviews what the tool produces, how the new hand-offs are coordinated and what capacity is needed to sustain it, where the work happens.

What to do about it

  • Look first at the use that already exists: who uses AI, for which tasks and with what results.
  • Move from individual use to team agreements: what can be done with AI, what gets reviewed and who is accountable.
  • Measure more than access: whether the way of working changed, and whether that change holds.

More in AI training. What a course does not build, in Measuring adoption beyond usage and in Geyto Learning.

Geyto Research publishes essays, evidence reviews and market notes on AI and technology adoption in organizations.

Frequently asked questions

How many Argentine SMEs use AI?

According to the nadIA survey (CEPE Di Tella and Fundar) of 402 SMEs with 10 to 249 employees, fielded in November and December 2025, 41.6% use at least one AI technology. Excluding software and IT services firms, 36.4%.

Why is AI adoption described as bottom-up?

Because people drive it more than organizations do. Among those who use AI at work, 85.6% learned on their own, 13.7% received training from their organization and 16% say there is an official strategy (CEPE Di Tella and Fundar, 2025).

What do companies lack to adopt AI?

Companies themselves point mainly to a lack of experience or knowledge (46.2%) and of qualified staff (42.2%), ahead of costs (29.2%). Most also have no usage policy and do not measure impact.

Sources

  1. nadIA, CEPE Di Tella and Fundar (2026). National survey on AI adoption in small and medium-sized enterprises in Argentina. Fieldwork by Fundación Observatorio PyME, November–December 2025. Summary in Perfil, 29 April 2026.
  2. Avenburg, A., Bendersky, A., Judzik, D. et al. (2025). Encuesta Nacional sobre Adopción de IA en Argentina. Relevamiento Individuos, August–September 2025. Public Policy Paper No. 37, CEPE, Universidad Torcuato Di Tella and Fundar.
  3. Taquion et al. (2026). Monitor Nacional de Inteligencia Artificial 2025. Summary in Inteligencia Argentina.
  4. Microsoft (2026). The state of global AI diffusion in 2026.
  5. Microsoft AI Diffusion Report, Q1 2026: figure for Argentina (21.9%). Summary in ITSitio, 18 June 2026.

Share