Generative AI advances rapidly, but its real impact on businesses stalls

Generative artificial intelligence expanded at unprecedented speed, but most business initiatives fail because organisations do not integrate it strategically.

Por El Medio Oriente
28 de agosto de 2026
A businessman in dark suit holds a transparent holographic screen displaying graphs, data and analysis indicators in blue and gold tones.
The graphs and data indicators on the holographic screen represent the tracking and analysis of business metrics, although many generative artificial intelligence initiatives in companies fail to achieve measurable results due to lack of strategic integration. (Perfil)
1 min de lectura
Tamaño del texto

Advanced generative artificial intelligence was adopted within weeks globally, with millions of people and organisations incorporating these tools to draft emails, process lengthy documents and carry out routine queries. However, behind the initial dazzle lies a paradox: while individual and recreational use expands rapidly, the real capacity to generate substantial improvements in organisational productivity remains stalled.

International consultancy Gartner projects that more than 70% of business initiatives in artificial intelligence and autonomous agents will fail by 2029. The main cause does not lie in algorithm limitations, but in the lack of internal order, the absence of prepared data and the inability to connect the technology with concrete problems in work processes, according to the consultancy.

In most companies and educational institutions, the technology is used as an isolated novelty. Licences are acquired or platforms are tested without altering existing methods, which produces a mirage of productivity: a sense of modernity that saves minutes in superficial tasks, but leaves intact the bottlenecks and structural inefficiencies.

To achieve measurable and sustainable impact, experts suggest structuring the incorporation of these solutions in four steps. First, understand the specific problem that needs to be solved through analysis of daily processes and identification of operational friction points. Second, work in interdisciplinary teams that bring together the professional who knows the process, the technical specialist and the management officer. Third, use instruction design techniques and query systems for internal databases to avoid fabricated answers. Fourth, conduct limited pilot tests before scaling the solution.

For small and medium-sized enterprises, as well as universities, this methodological approach opens a strategic opportunity to transform learning spaces into technology transfer centres, preparing professionals capable of turning technology into sustainable tools of high operational impact.

Generative AI advances but business impact stalls | El Medio Oriente