Skip to main content

A Snapshot of AI in Spanish SMEs and Professional Firms in 2026: 7 Lessons from the European Commission's Flash Eurobarometer 578

Analysis of official data for Spain (ES) compared to the EU-27 average (European Commission, DG EMPL, doi: 10.2767/3975834, 12,826 SMEs surveyed).

Map of Europe with a network of nodes, Spain highlighted in orange, three growing glass bars, and a balance scale
Quick answer: Analysis of official data for Spain (ES) compared to the EU-27 average (European Commission, DG EMPL, doi: 10.2767/3975834, 12,826 SMEs surveyed).

The European Commission (Directorate-General for Employment, Social Affairs and Inclusion, DG EMPL) has just published the Flash Eurobarometer 578 (Driving factors and obstacles related to the work environment for the adoption of Artificial Intelligence and Algorithmic Management by SMEs and mid-caps, catalogue KE-01-26-086-EN-N, doi: 10.2767/3975834), based on 12,826 interviews with companies employing between 1 and 749 people across the 27 Member States.

When comparing the specific data from Spain (ES) against the European average (EU27), seven key conclusions emerge for any professional firm, consultancy, SME, or public entity:

Official Comparative Table: Spain (ES) vs. European Union (EU27)

Official Indicator (Flash Eurobarometer 578)Spain (ES)EU-27 Average (EU27)Strategic Insights
Q1 · Current use of AI or automated systems42% (14% extensive + 28% limited)42 % (12 % + 30 %)Another 24% in Spain plans to adopt them (66% of the active market).
Q1d · Need for Business Adaptation75 % requires adaptation (28 % total + 47 % partial) and only 6 % accepts a standard system without adaptation58% adapted and 35% standard, not adaptedIn Spain, generic, decontextualized tools fail; professionals demand adaptation to their real procedures.
Q3 · Who makes the decisions82 % final human decision (50 % AI suggests but human decides + 32 % 100 % human) and only 1 % without human intervention90% human decision and 2% non-humanUndelegable human supervision and signature is the overwhelming cultural and operational standard.
Q4 · Lack of confidence in professional and personal decisions53% distrust (28% + 25%) versus 39% who trust49 % distrust vs 44 % trustTrust is not built with marketing: it is built with verifiable sources and traceability.
Q7f / Q7c / Q7a · Key drivers74% improve work organization, 73% improve service quality, and 67% reduce costs71 % organization, 75 % quality and 73 % costsIn Spain, the #1 driver is better organization of the daily workflow (74%).
Q8b and Q8c · The two major obstacles in Spain60% lack of internal expertise to integrate it into existing processes and 60% uncertainty regarding legal/regulatory standards48% lack of internal experience and 49% legal uncertaintySpain surpasses European concern by +12 and +11 points regarding how to integrate it into the process and with what legal certainty.
Q8d · Fear of increased surveillance38% identify it as a significant difficulty33% in the EU-27It demands separating work coordination and team handover from any punitive surveillance.
Q9 · Factor #1 for building team trust46% Staff training, 33% clear information on how decisions are made, and 26% human review/override35% training, 30% clear information, and 36% human reviewSpain leads the demand for real training (+11 points over the EU), in direct alignment with Article 4 of the European AI Regulation.

Conclusion from the Jaraba™ Method

Official data from the European Commission confirms what we observe daily in law firms, consultancies, and public administrations: the obstacle in Spain is not the rejection of technology, but rather the lack of a governed work system that integrates clear processes (Q7f: 74 %), real sectoral adaptation compared to generic solutions (Q1d: 75 % vs 6 %), regulatory rigor verified against official sources (Q8c: 60 %), final human decision (Q3: 82 %), and practical training for teams (Q9: 46 %).

Conversation 0

Has something similar happened to you, or do you see it differently? Share your thoughts here.

Comments are quickly reviewed before being published to prevent spam.