European enterprises are divided about adoption of AI for digital transformation, with two-in-five embracing AI technologies and two-in-five disinclined to engage with them. But the remainder are interested, and the diversity of AI technologies being adopted suggests decent momentum.
This is the conclusion of a new study carried out by market research firm Ipsos for the European Commission. It finds the main barriers to adoption are internal, with enterprises variously hindered because of a perceived lack of skills, high cost of adoption, and inflexibility of existing operational processes.
In general, the findings suggest European enterprises are powering ahead with ‘awareness’ of AI, as an orchestrating technology for digital transformation. The study results claim “almost universal” awareness, with 78 percent of enterprises familiar with the concept, notably of its role in anomaly detection and autonomous machines. Seven percent are ‘unaware’; 15 percent are ‘unsure’.
“Awareness of AI is clearly not a major barrier to the adoption of AI in Europe except amongst a small cohort of businesses,” the report states. As much as 42 percent of enterprises already use at least one type of AI technology, 25 percent use at least two types, and 18 per cent will adopt AI in the next two years.
But European enterprises show a significant divide at the ‘adoption’ stage, between ‘adopters’ (42 percent) and ‘non-adopters’ (40 percent). The remaining 18 percent have plans to adopt AI in the next two years.
Large businesses are more likely to be adopters compared to smaller businesses, and indeed to make more of it, considering their larger economies of scale and potential return on investment, the report says. Almost double the proportion of large enterprises (39 percent) use two or more AI technologies compared to micro-sized (21 percent) and small enterprises (22 percent).
Adoption ranges from three percent of enterprises using AI for ‘sentiment analysis’ to 13 percent for ‘anomaly detection’ and ‘process/equipment optimisation’, despite two in five adopting at least one of the 10 AI techniques listed by Ipsos in the survey. “Whilst uptake is relatively high amongst enterprises and differences in the adoption of specific technologies exist, there is no concentration of a specific technology,” the report says.
Equally, different sectors have different rates of adoption. The IT sector (63 percent) shows the highest rate. Outside of IT, differences in AI adoption across sectors is not very pronounced, says the study. “There are forerunners that experiment with all kinds of AI such as IT and the financial sectors. Conversely, sectors such as the construction sector are the lowest in terms of adopting different AI technologies, possibly because adoption is less relevant.”
It notes IT puts AI to use prominently for recommendation engines, and the financial sector employs it commonly to tackle wire fraud.
Three in five enterprises purchase AI software or ready-to-use systems, and two in five hire external providers to develop AI applications (59 and 38 percent); one in five, mostly large firms, develop AI technologies in house (20 percent), or use modified open source or commercial AI software (20 and 24 percent).
The study identifies three key internal barriers to AI adoption, around difficulties in hiring new staff with the right skills (57 percent), the cost of adoption (52 percent) and the cost of adapting operational processes (49 percent). External barriers include regulation, with 29 percent stating the need for new laws, and data standardisation, with 33 percent struggling strict standards for data exchange; 33 percent also worry about potential damages when it comes to adopting AI technologies.
The study concludes that exponential growth of AI in European enterprises is out-of-reach in the short-term, but that the diversity of adoption will continue at a fast pace. “The growth in the uptake is diversified across AI technologies and, depending on the barriers businesses currently face, is likely to result in a healthy growth reflective of the current uptake of these AI technologies,” it says.
