weeklyAI · Week of 27 September 2026weeklyAI · Semana del 27 de septiembre de 2026

← Your business← Su negocio

survey · Amfiteatru Economic · la publicación, 1 may 2026 · gratis

En Rumania casi nadie usa IA. Los que la adoptan por presión de la competencia, no por la herramienta.

Un sondeo a 285 pymes rumanas halló que lo que más pesa es mirar al rival y tener respaldo de la gerencia. Usar IA todavía no se tradujo en mejores resultados.

Versión breve · la versión detallada sigue, unos 6 min

Pregunte a weeklyAI

Pregúnteme por este estudio: a quiénes se estudió, qué encontró y qué no dice.

Las conversaciones se guardan mientras exista weeklyAI, para mejorar la publicación. Se responde en el idioma en que usted escribe.

El estudio, de un vistazo
Quiénes
Pequeñas y medianas empresas (pymes)
Cuántos
285 empresas
Dónde
Rumania
Cuándo
Entre octubre y diciembre de 2025
Tipo de estudio
survey
Quién lo hizo
Investigadores de la Bucharest University of Economic Studies, Kocaeli University y University of Bucharest
El límite que importa
Es un sondeo rumano de unos meses; no dice qué pasará en su mercado.
Tipo de apoyo que las pymes rumanas dicen necesitar para adoptar IA
Capacitación y desarrollo de habilidades37.9%
Regulaciones claras y simples20.4%
Subsidios financieros18.9%

Porcentaje de 285 pymes rumanas encuestadas que mencionaron cada tipo de apoyo; no suman 100% porque hubo otras respuestas.

Uso de inteligencia artificial: Rumania frente al conjunto de la Unión Europea

Rumaniafrente aUnión Europea

En Rumania solo el 3% de las empresas usaba IA, la cifra más baja de la UE; en el conjunto de la UE ese uso pasó del 8% en 2023 al 13.5% en 2024.

Ver · 0:57

En Rumania, apenas 3% de las empresas usa inteligencia artificial, el nivel más bajo de la Unión Europea. En el resto del bloque, el promedio subió de 8% a 13.5% entre 2023 y 2024. La mayoría de las pymes rumanas sigue sin estas herramientas.

Entre octubre y diciembre de 2025, un equipo de investigadores encuestó a 285 pymes de ese país. Entre las más destacadas hay empresas de servicios, manufactura, comercio y tecnología de la información. Preguntaron por su uso de tecnología, el respaldo de la gerencia, la presión de la competencia y sus resultados.

El patrón que apareció fue claro. Las empresas que sentían más presión de sus competidores percibían un mayor beneficio en adoptar IA, y lo mismo ocurría donde la gerencia respaldaba el cambio con objetivos, inversión y comunicación. Ese beneficio percibido, a su vez, iba de la mano con mejores resultados declarados.

Esto es un sondeo rumano, una foto de unos meses. No dice qué pasará en su mercado ni qué funcionará mañana. Rumania tiene un nivel de digitalización algo por debajo del promedio europeo, y eso limita cuánto se puede trasladar a otros países.

Usar IA, por sí solo, no mostró relación significativa con el desempeño. Tampoco el tamaño de la empresa. El respaldo gubernamental —cursos, capacitación, consultoría— sí se asoció directamente con mejores resultados, pero no a través del beneficio percibido. De cada diez empresas encuestadas, solo dos habían implementado soluciones de IA; tres no pensaban hacerlo.

Los resultados son lo que los propios encuestados reportaron, no cifras auditadas. Son percepciones, no balances.

Las plataformas más mencionadas fueron las de comunicación y colaboración, usadas por poco más de la mitad de los encuestados, seguidas por las de almacenamiento en la nube y, en tercer lugar, herramientas de IA como ChatGPT, Copilot o Gemini. Más de la mitad de las empresas usa entre una y tres plataformas; más de una cuarta parte, ninguna. Lo que más piden para avanzar es capacitación y habilidades, seguido por reglas claras y subsidios.

Para quien dirige una pyme en América Latina o Norteamérica, la lectura práctica es esta: observar a la competencia y conseguir que la dirección se comprometa de verdad parece pesar más, hoy, que apurarse a comprar herramientas. Antes de firmar con un proveedor, pregunte qué respaldo concreto ofrece su equipo directivo y qué hará usted si la herramienta no rinde.

¿Qué evidencia tiene usted de que su competidor está ganando por la IA, y no por otra cosa?

Qué significa para usted

Mientras tanto, fíjese en algo concreto: en el estudio, usar inteligencia artificial no se vinculó por sí solo con mejores resultados, y el tamaño de la empresa tampoco. Pregúntese qué respaldo real da su gerencia y qué haría usted si la herramienta no rinde, sin dar por hecho que comprarla bastará.

Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451

Quién pagó: Este trabajo fue apoyado por una subvención del Ministerio de Investigación, Innovación y Digitalización, CNCS–UEFISCDI, proyecto n.º PN-IV-P1-PCE-2023-0185, dentro del PNCDI IV; los autores agradecen a ISense Solutions su apoyo en la recogida de datos; el artículo no dice si los financiadores tuvieron alguna influencia.

Los hallazgos de otros estudios que aquí se mencionan los conocemos por este documento, que fue el que leímos; no abrimos cada uno de esos estudios.

Versión detalladaLos pasajes copiados del artículo, las ilustraciones y cada fuente con cuánto leímos de ella · 1127 palabras · unos 6 minLeerla →Cerrar

Lo que la presión de la competencia mueve, y lo que no, en las pymes que prueban la IA

Un sondeo a 285 pymes rumanas encontró que la ventaja que los dueños perciben viene del apuro por no quedar atrás y del respaldo de la gerencia, no del uso de la herramienta en sí.

Así podría verse · ilustración generada por weeklyAI.watch, no es una fotografía

En una encuesta a 285 pequeñas y medianas empresas de Rumania, hecha entre octubre y diciembre de 2025, los investigadores encontraron que la presión de la competencia fue el factor que más se relacionó con la ventaja que los dueños y gerentes percibían al adoptar tecnologías nuevas1. El respaldo de la gerencia apareció como el segundo factor fuerte en esa misma dirección2. El tamaño de la empresa tuvo un efecto leve y de signo contrario: las más grandes percibían un poco menos de ventaja3. Y el uso real de inteligencia artificial no alcanzó a ser un predictor significativo de esa ventaja percibida4.

Sobre el desempeño del negocio, medido por lo que los propios encuestados dijeron sobre ventas, costos y relaciones, la ventaja percibida sí apareció como el eslabón central5. El respaldo de la gerencia mejoró el desempeño a través de esa ventaja percibida5. La presión de la competencia también se relacionó con el desempeño a través de esa ventaja percibida5. El apoyo del gobierno tuvo un aporte directo e independiente sobre el desempeño6. Ni el tamaño de la empresa ni el uso de IA mostraron efectos significativos sobre el desempeño6. El propio artículo advierte que el uso de IA todavía no había rendido beneficios tangibles comparables entre las empresas estudiadas7.

Conviene entender qué se midió y qué no. Esto fue un sondeo con cuestionario a 285 empresas8. Cuando la inteligencia artificial aparece en la lista de tecnologías, se trata de capacidades concretas: automatizar procesos, hacer análisis predictivo, optimizar operaciones y personalizar el servicio9. Eso es lo que un dueño de pyme tendría en mente al responder si la herramienta le sirvió o no.

El punto de partida del país donde se hizo el estudio explica por qué el uso de IA no movió la aguja. En Rumania solo el 3% de las empresas usaba inteligencia artificial, la cifra más baja de la Unión Europea10. En el conjunto de la UE, ese uso pasó del 8% en 2023 al 13.5% en 202410. Y en digitalización básica, el 73% de las pymes europeas había alcanzado un nivel mínimo, frente al 69% de las rumanas11. El propio artículo señala que la etapa temprana de aplicación de la IA dificultó identificar beneficios tangibles comparables entre las empresas estudiadas7.

Así podría verse · ilustración generada por weeklyAI.watch, no es una fotografía

Vale la pena mirar estudios parecidos hechos en otros países, aunque de estos solo pudimos leer el resumen, no el texto completo. Un trabajo polaco, con un sondeo a 112 pymes y entrevistas a 13 gerentes, encontró que el uso de IA se estaba volviendo común pero desigual y sobre todo operativo, con los efectos más fuertes en reducción de carga de trabajo y ahorro de tiempo, especialmente en empresas de servicios donde la herramienta se usa con intensidad12. Ese mismo trabajo advierte que esos efectos no deben leerse como reducción directa de costos: la IA mejora la productividad y la capacidad de trabajo, pero crea costos nuevos en herramientas pagadas, preparación de datos, integración, verificación de resultados y gobierno de la herramienta13. Y agrega que el paso del uso improvisado a una implementación estratégica depende menos del tamaño de la empresa que de la madurez de sus procesos, sus capacidades y la calidad de sus datos14.

Otro estudio, esta vez en empresas de un país distinto, encontró que las aplicaciones de IA tuvieron un impacto positivo y estadísticamente significativo en la productividad laboral y en la reducción de costos operativos, aunque su efecto sobre la eficiencia de los procesos de negocio fue moderado y parcialmente limitado1516. La aplicación operativa, como automatización y análisis de datos, resultó el factor más importante de los efectos económicos17. Ese trabajo también señala que el simple aumento de aplicaciones de IA con el tiempo no lleva necesariamente a más eficiencia sin una implementación concreta y dirigida18.

Sobre lo que los propios dueños y gerentes rumanos dicen necesitar, el estudio recogió respuestas reveladoras: el 37.9% pidió capacitación y desarrollo de habilidades, el 20.4% pidió regulaciones claras y simples, y el 18.9% consideró útiles los subsidios financieros19. Entre las respuestas, la capacitación y el desarrollo de habilidades fue la más mencionada, seguida de regulaciones claras y simples, y de subsidios financieros19. El trabajo polaco del que solo leímos el resumen apunta en la misma dirección: los obstáculos cambian con la madurez, y mientras las empresas que recién empiezan enfrentan falta de conocimiento, tiempo y casos de uso claros, las que ya avanzaron chocan con problemas de calidad de datos, resultados inventados por el sistema, seguridad, integración y gobierno20.

Así podría verse · ilustración generada por weeklyAI.watch, no es una fotografía

Así lo leemos nosotros. El patrón que aparece aquí es viejo y conocido: cuando una tecnología se mueve rápido, la protección que venía del oficio aprendido, del puesto fijo o del tamaño de la empresa se debilita, y cada negocio queda negociando solo sus condiciones en un mercado que cambia sin reglas claras. Lo que este estudio sugiere, sin probarlo, es que en las pymes de nuestra región los dueños y gerentes van a sentir la presión de adoptar IA como una forma de no quedar expuestos, aunque todavía no tengan claro qué reglas ni qué protecciones aplican a sus propios empleados y datos. Si nos equivocamos, lo veremos en pymes que adoptan IA y mantienen o mejoran las condiciones de su gente y sus márgenes sin necesidad de renegociar todo individualmente. Lo que usted puede hacer con esto, ya: antes de comprar cualquier herramienta, revise qué pasa con los contratos, las tareas y los datos de su gente si la herramienta cambia de precio o de dueño; y pregunte a sus empleados qué tareas temen perder y qué necesitan aprender.

También leímos otros trabajos que apuntan en dirección contraria a este, y conviene decirlo. Estas diferencias no invalidan el resultado rumano: lo que muestran es que un hallazgo de este tipo todavía no ha demostrado trasladarse a otros países, otras poblaciones u otros sistemas, y que la evidencia sobre el tema sigue siendo contradictoria. Tenga esto presente antes de sacar conclusiones tajantes sobre lo que la IA hará o no hará por un negocio como el suyo.

Nuestra lectura de todo esto, en una frase que usted puede usar mañana: la presión de la competencia y el respaldo de la gerencia son las palancas que la evidencia sostiene; el uso de la herramienta por sí solo, todavía no. Pregúntese, antes de firmar cualquier compra o de anunciar cualquier cambio: ¿esto mejora un proceso concreto y medible, o solo me hace sentir que no estoy quedando atrás?

De dónde sale cada dato de contexto, y cuánto leímos de cada documento

  1. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Competitive pressure (CP) strongly, positively and significantly influences perceived relative advantage (β = 0.537, p <.001), meaning that SMEs facing higher competitive pressure perceive a greater relative advantage in adopting new technologies, such as AI platforms and other digital platforms."
  2. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Top Management Support (TMS) is another factor with a strong, positive and significant influence on perceived relative advantage (β = 0.425, p <.001)."
  3. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The results show that perceived relative advantage is mainly determined by competitive pressure and support from company management, while company size has a modest negative influence. Government support and AI use are not significant predictors."
  4. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The use of artificial intelligence (UAI), although it has a positive influence (β = 0.122), is not a significant predictor of perceived relative advantage (p = 0.105)."
  5. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Furthermore, perceived relative advantage is a key mediator linking strategic and organisational factors to performance. It fully mediates the relationship between top management support for AI adoption and business performance and partially mediates the relationship between competitive pressure and performance."
  6. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Government support has a direct and independent contribution to business performance, while company size and AI use do not have significant effects on performance at present."
  7. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The limitations of the research stem from the relatively early stage of application of AI and other innovative technologies among SMEs in Romania, which has made it difficult to identify commensurable, tangible benefits at present among certain companies surveyed."
  8. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "This paper aims to understand how artificial intelligence (AI) is integrated into the business processes of SMEs by conducting quantitative marketing research on a sample of 285 companies operating in different regions of the country."
  9. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The main capabilities of AI technology are: process automation (Dalsaniya and Patel, 2022), predictive analytics (Basal, Moulai and Cetin, 2025), operations optimisation (Martins, 2024) and service personalisation (Stancu et al., 2023; Cherukuri, 2024)."
  10. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The use of artificial intelligence (AI) among businesses in the European Union has increased from 8% in 2023 to 13.5% in 2024, highlighting a rapid trend towards the adoption of digital technologies. In Romania, however, this use is much lower, at only 3%, placing the country last in the European ranking (Eurostat, 2025)."
  11. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Eurostat reports (2025) show that 73% of SMEs in the EU have achieved a basic level of digitisation, while in Romania only 69% have reached this minimum level, which highlights a level of digitisation slightly below the European average."
  12. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "The results show that AI adoption in SMEs is increasingly common but remains uneven and mostly operational. The strongest effects concern workload reduction and time efficiency, particularly in service firms and where AI is used intensively."
  13. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Rather, AI improves productivity and work capacity while creating new costs related to paid tools, data preparation, integration, output verification, and governance."
  14. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "The transition from ad hoc use to strategic implementation depends less on firm size alone and more on process maturity, capabilities, and data readiness."
  15. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "The research results show that different applications of artificial intelligence have a statistically significant, positive impact on labor productivity and on reducing operating costs."
  16. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "In contrast, their impact on business process efficiency is moderate and partially limited."
  17. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "The operational application of artificial intelligence, such as automation and data analysis, has proven to be the most important factor in economic effects."
  18. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "On the other hand, the mere growth of AI applications over time does not necessarily lead to increased efficiency without targeted and concrete implementation."
  19. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Regarding the type of support needed by SMEs to adopt AI, 37.9% of respondents indicate training and skills development, 20.4% feel the need for clear and simple regulations, and 18.9% consider financial subsidies useful."
  20. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Barriers also change with maturity: early-stage firms face a lack of knowledge, time, and clear use cases, whereas advanced users encounter data quality, hallucinations, security, integration, and governance problems."

Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451

Quién pagó: Este trabajo fue apoyado por una subvención del Ministerio de Investigación, Innovación y Digitalización, CNCS–UEFISCDI, proyecto n.º PN-IV-P1-PCE-2023-0185, dentro del PNCDI IV; los autores agradecen a ISense Solutions su apoyo en la recogida de datos; el artículo no dice si los financiadores tuvieron alguna influencia.

Los hallazgos de otros estudios que aquí se mencionan los conocemos por este documento, que fue el que leímos; no abrimos cada uno de esos estudios.

survey · Amfiteatru Economic · the paper, 1 May 2026 · free

Competitors at the Door, Not AI, Moved These Businesses

In a survey of 285 small firms in Romania, pressure from rivals and backing from the boss predicted whether managers saw AI as worth it. Actually using AI did not.

Short version · the longer version follows, about 7 min

Ask weeklyAI

Ask me about this study: who was studied, what it found, and what it does not say.

Conversations are saved for as long as weeklyAI exists, to improve the publication. Answers come in the language you write in.

The study at a glance
Who
owners, managers and employees in small and medium-sized businesses
How many
285 companies
Where
Romania
When
October to December 2025
Kind of study
survey
Who did it
Bucharest University of Economic Studies, Kocaeli University, University of Bucharest
The limit that matters
Early stage of AI use in Romania made tangible benefits hard to identify
What support small firms say they need to adopt AI
Training and skills development37.9%
Clear and simple regulations20.4%
Financial subsidies18.9%

Shares of surveyed Romanian small and medium businesses naming each kind of support; the three do not add to 100 because other options were also offered.

So if your competitors are already using AI, that alone may be the strongest signal you have.
weeklyAI's reading
Watch · 0:53

In Romania, only about 3 in 100 businesses use artificial intelligence — the lowest rate in the European Union. Among the small and medium-sized companies researchers surveyed there, most still run without AI tools. A little over a quarter used none of the digital platforms the questionnaire asked about at all.

The researchers wanted to know what makes a small firm see a new technology as an advantage. So they asked 285 owners, managers and employees in Romanian small and medium businesses, between October and December 2025, about their competitors, their bosses, their tools and their results.

Companies that felt the heat from competitors were the most likely to see AI as an edge. So were companies where senior management actively pushed for it — setting goals, committing money, talking it up. Firm size pointed the other way, but only slightly: bigger companies saw a bit less advantage.

The study is from Romania, a snapshot taken in late 2025, and it is a small one. Where only about 3 in 100 businesses use AI, the picture can look very different from markets where the tools are already common. The authors themselves say the link between using AI and performing better could change as experience grows.

Actually using AI made no measurable difference to performance — not directly, and not through the perceived advantage. Government support, by contrast, was linked to better performance directly, in the form of revenue, cost and efficiency indicators. But it did not work by changing how managers saw AI's advantages.

What the study really measures is what people believed, not audited books. Performance here means managers' own reports of customer relations, efficiency, sales and costs.

Larger firms did not perform better because of their size — not directly, not through perceived advantage. In this sample, a small company that moved on new technology was not automatically behind a bigger one.

So if your competitors are already using AI, that alone may be the strongest signal you have. And if your own management is not behind the push, the perceived upside may never show up.

Ask this before your next technology purchase: who is pushing for it — the market, or the people who will have to make it work?

What this means for you

For your own small business, the finding is worth holding lightly: competitors pushing and management backing were linked to seeing AI as an edge, not to actually performing better. Since AI use showed no measurable effect, watch whether your own team sees the advantage before spending, and ask who is pushing the purchase.

Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451

Who paid: This work was supported by a grant from the Ministry of Research, Innovation and Digitalisation, CNCS–UEFISCDI, project no. PN-IV-P1-PCE-2023-0185, within PNCDI IV; the authors thank ISense Solutions for support in data collection; the article does not say whether funders had any say.

The findings of other studies mentioned here are known to us through this document, which is the one we read; we did not open each of those studies.

The longer versionThe passages copied from the paper, the pictures, and every source with how much of it we read · 1484 words · about 7 minRead it →Close

Competitive Pressure and Management Backing, Not AI Use Itself, Linked to Romanian SMEs' Sense of Advantage

A survey of 285 firms in Romania finds that wanting to keep up and having the boss behind the effort mattered more than the tools already in place.

How it could look · illustration generated by weeklyAI.watch, not a photograph

The study set out to understand how artificial intelligence is integrated into the business processes of small and medium-sized enterprises, by conducting quantitative marketing research on a sample of 285 companies operating in different regions of Romania1. Researchers surveyed employees and representatives with a decision-making or executive role, collecting data between October and December 2025 and analysing it with a technique that tests how well the questions group together into the factors the authors had in mind.

What did they find? Perceived relative advantage — the sense that a new technology is better than what you already have — was mainly determined by competitive pressure and support from company management, while company size had a modest negative influence2. Government support and actual AI use were not significant predictors of that sense of advantage2.

The numbers behind those two main drivers: competitive pressure showed a strong, positive link to perceived relative advantage, meaning that SMEs facing higher competitive pressure perceive a greater relative advantage in adopting new technologies such as AI platforms and other digital platforms3. Top management support was another factor with a strong, positive link to that same perception4.

On performance, perceived relative advantage had a direct effect on business performance, and the authors describe it as a key mediator linking strategic and organisational factors to performance — it fully mediates the relationship between top management support for AI adoption and business performance, and partially mediates the relationship between competitive pressure and performance5. Government support had a direct and independent contribution to business performance, while company size and AI use did not have significant effects on performance at present6.

Here is where a reader should slow down. The use of artificial intelligence, although it had a positive influence, was not a significant predictor of perceived relative advantage7. And AI use had no significant direct or indirect effects on business performance8. In plain terms: the study found no significant link between using AI and business performance.

The authors are direct about why. The limitations of the research stem from the relatively early stage of application of AI and other innovative technologies among SMEs in Romania, which made it difficult to identify commensurable, tangible benefits at present among certain companies surveyed9. They also note that replicating the research methodology through longitudinal studies, applied in the future to the same panel of companies, could lead to a positive and significant change in the nature of the relationship between AI use and business performance10. In other words: ask again later.

To picture how small this starting point is: Eurostat reports show that 73% of SMEs in the EU have achieved a basic level of digitisation, while in Romania only 69% have reached this minimum level11. On AI specifically, use among EU businesses rose from 8% in 2023 to 13.5% in 2024 — and in Romania it sits at only 3%, placing the country last in the European ranking12. Picture a room of a hundred business owners: roughly three have touched AI at all.

How it could look · illustration generated by weeklyAI.watch, not a photograph

What counts as AI here is not exotic. The main capabilities are process automation, predictive analytics, operations optimisation and service personalisation13 — that is, taking repetitive work off a person's desk, guessing what comes next from past patterns, smoothing how work flows, and tailoring what a customer receives.

The Romanian sample is not a tech sector. Among the 285 firms, the most prominent are services at 30%, manufacturing at 14%, trade at 11.6% and IT&C at 6%14. When the study asked what support these firms would need to adopt AI, 37.9% of respondents indicated training and skills development, 20.4% felt the need for clear and simple regulations, and 18.9% considered financial subsidies useful15. Notice the order: skills first, rules second, money third.

Other work we read — in these cases only the summaries, with the full papers behind subscriptions — complicates the picture. According to the summary of a study of 112 SMEs in Poland's Kuyavian–Pomeranian region, including 70 AI-using firms, with 13 in-depth manager interviews, AI adoption in SMEs is increasingly common but remains uneven and mostly operational, with the strongest effects on workload reduction and time efficiency, particularly in service firms and where AI is used intensively16. The same summary reports that advanced adoption raises the probability of perceiving workload and cost-related effects, but that these should not be read simply as direct cost reduction — rather, AI improves productivity and work capacity while creating new costs for paid tools, data preparation, integration, output verification and governance1718. It also describes a staged path: from curiosity-driven experimentation, through augmenting cognitive work, to workflow integration and, in selected cases, business model innovation19. And it says the move from ad hoc use to strategic implementation depends less on firm size alone and more on process maturity, capabilities and data readiness20.

A separate summary of a study of 228 SMEs, analysed with multiple linear regression, reports that different applications of AI have a statistically significant, positive impact on labour productivity and on reducing operating costs, while their impact on business process efficiency is moderate and partially limited2122. That summary also finds the operational application of AI — automation and data analysis — to be the most important factor in economic effects, with its use in managerial decision-making significant but somewhat weaker2324 — and warns that the mere growth of AI applications over time does not necessarily lead to increased efficiency without targeted and concrete implementation25.

How it could look · illustration generated by weeklyAI.watch, not a photograph

A third summary, of an exploratory multiple-case study built on semi-structured interviews with fourteen women leading established SMEs in Chile, describes four adoption profiles — non-adopter, AI experimenter, substantive AI adopter and integrated AI adopter — and reports that participants used AI primarily to support customer relationship management, while describing limited AI readiness as a constraint tied to learning needs2627. That summary is careful about what it can claim: instead of identifying causal determinants of successful adoption, it highlights recurring patterns and relationships that may inform future theory development and comparative research28.

Here is how we read it. The pattern in this Romanian survey is a familiar one in business: people copy what a peer they respect is doing, not because they have worked out the reasons, but because they want to stand where that peer stands. The study found competitive pressure to be the strongest driver of the sense that AI is worth having — and yet AI use itself was not tied to better performance. Put those together and the pressure may be coming more from watching rivals than from a clear internal case. We could read only the summaries of those papers; the full texts sat behind subscriptions. So read the Romanian result as one country's snapshot, not as a verdict on your own.

What does that lead us to expect where you work? When you hear that competitors are using AI, the useful move is to ask who exactly is using it and what they get from it before you spend — separate the wish to keep up from a tested reason to adopt. We would also expect that a tool can become the default not because it is best but because everyone around you is already on it and leaving feels costly; this study did not measure that, so treat it as something to check, not something it proved. You would know we are wrong if firms like yours could switch AI or cloud providers easily and cheaply, or if they chose suppliers mainly on measured results. Before you commit, ask how hard it would be to move your data and work elsewhere, and keep a copy you control; watch for contracts that make leaving expensive. And ask the person who runs your current systems what would break if you changed them — plan the change in small steps, with a way back, rather than treating their caution as obstruction.

We also read, from other work, that this kind of system carries a risk worth naming plainly: early-stage firms tend to face a lack of knowledge, time and clear use cases, while more advanced users run into data quality, invented answers, security, integration and governance problems. That is a warning about what to watch for, not a finding of this study. The practical version: pick one process, tie the tool to a measurable outcome on that process, and invest in the literacy and the rules around it before scaling. And when a vendor or a rival tells you the moment has arrived, ask what changed in the process — not just what changed in the software.

What did the tool actually change in your process — and could you name the number?

Where each piece of context comes from, and how much of it we read

  1. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "This paper aims to understand how artificial intelligence (AI) is integrated into the business processes of SMEs by conducting quantitative marketing research on a sample of 285 companies operating in different regions of the country."
  2. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The results show that perceived relative advantage is mainly determined by competitive pressure and support from company management, while company size has a modest negative influence. Government support and AI use are not significant predictors."
  3. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Competitive pressure (CP) strongly, positively and significantly influences perceived relative advantage (β = 0.537, p <.001), meaning that SMEs facing higher competitive pressure perceive a greater relative advantage in adopting new technologies, such as AI platforms and other digital platforms."
  4. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Top Management Support (TMS) is another factor with a strong, positive and significant influence on perceived relative advantage (β = 0.425, p <.001)."
  5. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Furthermore, perceived relative advantage is a key mediator linking strategic and organisational factors to performance. It fully mediates the relationship between top management support for AI adoption and business performance and partially mediates the relationship between competitive pressure and performance."
  6. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Government support has a direct and independent contribution to business performance, while company size and AI use do not have significant effects on performance at present."
  7. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The use of artificial intelligence (UAI), although it has a positive influence (β = 0.122), is not a significant predictor of perceived relative advantage (p = 0.105)."
  8. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The use of artificial intelligence (UAI) has no significant direct (β = -0.089, p = 0.283) or indirect (β = 0.067, p = 0.245) effects on business performance."
  9. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The limitations of the research stem from the relatively early stage of application of AI and other innovative technologies among SMEs in Romania, which has made it difficult to identify commensurable, tangible benefits at present among certain companies surveyed."
  10. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Replicating the research methodology by conducting longitudinal studies, to be applied in the future to the same panel of companies, could lead to a positive and significant change in the nature of the relationship between AI use and business performance."
  11. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Eurostat reports (2025) show that 73% of SMEs in the EU have achieved a basic level of digitisation, while in Romania only 69% have reached this minimum level, which highlights a level of digitisation slightly below the European average."
  12. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The use of artificial intelligence (AI) among businesses in the European Union has increased from 8% in 2023 to 13.5% in 2024, highlighting a rapid trend towards the adoption of digital technologies. In Romania, however, this use is much lower, at only 3%, placing the country last in the European ranking (Eurostat, 2025)."
  13. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The main capabilities of AI technology are: process automation (Dalsaniya and Patel, 2022), predictive analytics (Basal, Moulai and Cetin, 2025), operations optimisation (Martins, 2024) and service personalisation (Stancu et al., 2023; Cherukuri, 2024)."
  14. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "The sample consists of 285 SMEs active on the Romanian market. Among these, the most prominent are those in services (30%), manufacturing (14%), trade (11.6%) and IT&C (6%)."
  15. Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451 - the article this story is about — the whole article — the passage: "Regarding the type of support needed by SMEs to adopt AI, 37.9% of respondents indicate training and skills development, 20.4% feel the need for clear and simple regulations, and 18.9% consider financial subsidies useful."
  16. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — only the abstract - the full text could not be fetched — the passage: "The results show that AI adoption in SMEs is increasingly common but remains uneven and mostly operational. The strongest effects concern workload reduction and time efficiency, particularly in service firms and where AI is used intensively."
  17. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — only the abstract - the full text could not be fetched — the passage: "Advanced AI adoption increases the probability of perceiving workload and cost-related effects. However, these effects should not be interpreted simply as direct cost reduction."
  18. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — only the abstract - the full text could not be fetched — the passage: "Rather, AI improves productivity and work capacity while creating new costs related to paid tools, data preparation, integration, output verification, and governance."
  19. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — only the abstract - the full text could not be fetched — the passage: "The interviews show that AI implementation follows a staged path: from curiosity-driven experimentation, through cognitive work augmentation, to workflow integration and, in selected cases, AI-enabled business model innovation."
  20. Polasik M, Czarkowska M, Śniadkowski W, Bagniewski B, Meler A. (2026). The Impact of the Implementation of the AI Systems in Small and Medium Enterprises in Poland: Scale of Usage, Productivity, and Unperceived Sustainability. Sustainability. https://doi.org/10.3390/su18136503 — only the abstract - the full text could not be fetched — the passage: "The transition from ad hoc use to strategic implementation depends less on firm size alone and more on process maturity, capabilities, and data readiness."
  21. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — only the abstract - the full text could not be fetched — the passage: "The research results show that different applications of artificial intelligence have a statistically significant, positive impact on labor productivity and on reducing operating costs."
  22. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — only the abstract - the full text could not be fetched — the passage: "In contrast, their impact on business process efficiency is moderate and partially limited."
  23. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — only the abstract - the full text could not be fetched — the passage: "The operational application of artificial intelligence, such as automation and data analysis, has proven to be the most important factor in economic effects."
  24. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — only the abstract - the full text could not be fetched — the passage: "At the same time, its application in managerial decision-making also has a significant, but somewhat weaker impact."
  25. Bolfek M, Rajko M, Bolfek B. (2026). The Economic Effects of Artificial Intelligence Adoption in Small and Medium-Sized Enterprises. World. https://doi.org/10.3390/world7060103 — only the abstract - the full text could not be fetched — the passage: "On the other hand, the mere growth of AI applications over time does not necessarily lead to increased efficiency without targeted and concrete implementation."
  26. Ireta-Sánchez JM, Berkatsashvili M. (2026). Artificial Intelligence Adoption in Established Women-Led SMEs: A Multiple-Case Study of Competencies, Barriers, and Business Process Applications. Systems. https://doi.org/10.3390/systems14091120 — only the abstract - the full text could not be fetched — the passage: "Participants reported utilising AI primarily to support customer relationship management (CRM), while also describing limited AI readiness as a constraint associated with learning needs and experiences of AI adoption."
  27. Ireta-Sánchez JM, Berkatsashvili M. (2026). Artificial Intelligence Adoption in Established Women-Led SMEs: A Multiple-Case Study of Competencies, Barriers, and Business Process Applications. Systems. https://doi.org/10.3390/systems14091120 — only the abstract - the full text could not be fetched — the passage: "To distinguish between AI adoption and broader digitalisation and automation, the study defines the following adoption profiles: non-adopter, AI experimenter, substantive AI adopter, and integrated AI adopter."
  28. Ireta-Sánchez JM, Berkatsashvili M. (2026). Artificial Intelligence Adoption in Established Women-Led SMEs: A Multiple-Case Study of Competencies, Barriers, and Business Process Applications. Systems. https://doi.org/10.3390/systems14091120 — only the abstract - the full text could not be fetched — the passage: "Instead of identifying causal determinants of successful adoption, the findings highlight recurring patterns and relationships that may inform future theory development and comparative research on AI adoption in entrepreneurial contexts."

Filip, A., Stancu, A., Alniacik, U. et al. (2026). Adoption of Artificial Intelligence and SME Performance in Digital Ecosystems. Amfiteatru Economic. https://doi.org/10.24818/ea/2026/72/451

Who paid: This work was supported by a grant from the Ministry of Research, Innovation and Digitalisation, CNCS–UEFISCDI, project no. PN-IV-P1-PCE-2023-0185, within PNCDI IV; the authors thank ISense Solutions for support in data collection; the article does not say whether funders had any say.

The findings of other studies mentioned here are known to us through this document, which is the one we read; we did not open each of those studies.