European companies are at risk of falling behind in the artificial intelligence (AI) race, according to a new analysis that says they are being dragged down by years of underinvestment and outdated systems.
The report from the American technology company Cisco found that many European organisations lack the technical and financial foundations to scale up AI effectively.
It surveyed 8,000 senior information technology (IT) and business leaders responsible for AI strategy at organisations with over 500 employees across 26 industries. Nearly 1,500 were respondents from France, Germany, Italy, the Netherlands, Poland, Spain, and Sweden.
The report frames “AI infrastructure debt” as a modern offshoot of technical debt, a term coined in the early 1990s to describe how short-term fixes and postponed upgrades in software can create costly inefficiencies over time.
In much the same way, Cisco suggested that delayed investment and outdated infrastructure are already holding back AI’s growth. It highlighted rising workloads, limited compute capacity needed for AI, fragmented data systems, and inflexible networks as key challenges.
Across the seven European Union countries surveyed, 45 per cent of respondents said they expect AI workloads to grow by more than 30 per cent within three years.
However, only 23 percent said they have robust graphics processing unit (GPU) capacity and 66 per cent struggle to centralise their data. Another36 per cent said their networks cannot keep up with AI’s demands.
Only 9 per cent described their networks as flexible or adaptable.
AI readiness in Europe falls behind
Cisco said “these early warning signs point to a gap between AI ambition and operational readiness”.
While nearly half of EU respondents said their organisations have an AI roadmap, only 18 per cent make AI their top investment priority, compared with 79 per cent of the top performers globally. Top investors tend to see better performance, Cisco said.
The report also highlighted concerns around security. Only a third of European respondents say they are highly aware of AI-specific threats, and just a quarter have built AI into their wider security systems.
