Skip to content
Marsanne Studio
← The AI watch

Data study ·

84% of CIOs have cancelled an AI project because of legacy systems

Original title, in Spanish: “El 84% de los CIO cancela un proyecto de IA por el legacy”

Publication
Revista Byte TI (Spain)
Author
Alfonso Casas
Published on
1 October 2026 · read in the page structured data

Summary

According to a GFT Technologies study carried out by Wakefield Research among 945 CIOs and CTOs in 19 countries, at companies with annual revenue of at least $500 million, outdated infrastructure is a major brake on AI adoption. 84% of respondents say they have cancelled a pilot or an AI project because of the limitations of their legacy systems, and 93% believe that deploying AI without modernising infrastructure will cause a security crisis. The study also finds that 89% of CIOs fear an AI bubble, that 99% consider it important not to depend on a single vendor following the dispute between the US government and Anthropic, and that 51% have hired employees to correct work generated by AI.

Summary based on the full text of the article.

Read the article on Revista Byte TI →

Figures cited

  • 84 %

    CIOs who have cancelled an AI project because of legacy systems

    945 CIOs/CTOs in 19 countries, companies with annual revenue of $500M or more, August 2026

    « El 84% afirma que las limitaciones de sus sistemas legacy han provocado que su organización cancele un piloto o un proyecto de IA. »

    Source: GFT Technologies / Wakefield Research

  • 93 %

    CIOs fearing an AI-related security crisis on outdated infrastructure

    945 CIOs/CTOs in 19 countries

    « el 93% cree que no modernizar antes de ejecutar IA sobre infraestructura obsoleta acabará desencadenando una una crisis de seguridad que afecte a toda la organización. »

    Source: GFT Technologies

  • 89 %

    CIOs concerned about a possible AI bubble

    945 CIOs/CTOs in 19 countries

    « El 89% está preocupado por que la inversión global en IA pueda estar creciendo más rápido que el valor empresarial que puede generar de forma realista. »

    Source: GFT Technologies

  • 99 %

    Importance of not depending on a single AI vendor

    945 CIOs/CTOs in 19 countries

    « El 99% afirma que las posibles restricciones gubernamentales al acceso a la IA incrementan la importancia de no depender de un único proveedor de IA »

    Source: GFT Technologies

  • 42 %

    CIOs favouring in-house construction of AI infrastructure

    945 CIOs/CTOs in 19 countries

    « el 42% de los encuestados se inclina ahora por construir internamente la infraestructura de IA en lugar de adquirirla a proveedores externos. »

    Source: GFT Technologies

  • 51 %

    Companies that have hired staff to correct AI's work

    945 CIOs/CTOs in 19 countries

    « Más de la mitad (51 %) afirmó haber contratado a empleados específicamente para revisar o corregir el trabajo generado por la IA »

    Source: GFT Technologies

  • 26 %

    Companies that have rehired employees previously laid off

    945 CIOs/CTOs in 19 countries

    « el 26 % señaló haber readmitido a empleados a los que habían despedido anteriormente. »

    Source: GFT Technologies

Classification

Country
Worldwide scope
Company size
Large companies, Mid-sized companies
Industries
All industries, or industry not specified · Software vendors, IT services companies, telecoms
Companies
Wakefield Research (cited), GFT Technologies (study author), Anthropic (cited)
Keywords
legacy systems, cios, outdated infrastructure, ai bubble, it security, employment

Entry written by Marsanne Studio from the cited article, with the help of an AI model (claude-sonnet-5). The rights to the article and its illustrations belong to Revista Byte TI and its author. Entry translated from French by Claude Sonnet 5.5.