Artificial intelligence: good recipes from IDC to boost your projects
French companies display high digital maturity in Europe, with strong adoption of the cloud and a strategy focused on data management. Despite this, the adoption of artificial intelligence in France lags significantly behind other European countries.
This is the finding of an IDC report (sponsored by Microsoft) signed by Cyrille Chausson, Research Manager, IDC Europe.
“60% of French companies (compared to 51% in Europe) describe themselves as companies whose economic model is based on digital or which were built naturally with and on the cloud (“cloud-native”),” he explains. -he. But “they [ne] are today [que] 48% say they use AI, whatever the use case, compared to almost 68% for companies in Europe.”
This delay is also observed in the adoption of generative AI.
A more cautious, transversal and pragmatic French approach to AI
This caution could be partly explained by the desire to identify relevant use cases.
“Almost one in two French companies has structured their strategy by connecting AI to real expectations in terms of business results,” confirms Cyrille Chausson. Innovation, improving the customer experience and agility are the main targeted objectives.
French companies would therefore favor a pragmatic, but also transversal, approach. French organizations “only 28% place this budget in the sole hands of the IT department, compared to 42% in Europe”, figures the analyst.
“This approach with strong IT/business proximity allows French companies to have an agile approach […] and to modify the budgetary structure according to the progress of AI in the company and the business gains provided,” continues Cyrille Chausson.
41% of companies in France would plan to have this approach in the next two years, while 33% have already managed their AI budgets in this way in France.
A delay also in responsible AI
The fact remains that local companies would show another delay: in the governance of responsible AI. “Only 26% have built a governance committee, while one in three companies have already set it up in Europe.”
At the same time, this governance is seen as the best way to adapt to regulatory constraints. This is another major sticking point. “Regulations are slowing down the progress of AI in France,” writes Cyrille Chausson.
“Regulations are slowing down the progression of AI in France. »
Cyrille ChaussonIDC Europe
“Waiting for European or national regulations, linked in particular to digital sovereignty, or due to a lack of understanding of the constraints imposed by these laws, French companies are delaying their implementation of AI,” he notes. “However, the long-awaited European AI Act regulation, the text of which was published in December 2023, should help set the rules for the use of data. It should further structure companies’ strategies, reassuring them about their integration of AI into their business processes,” he adds.
Costs, lack of skills and training
But the main obstacles remain costs, lack of skills, and lack of training of employees in AI.
However, “specialized personnel (such as “prompt engineers” or “AI modelers”) and competent and trained employees [sont] two essential components for the propagation of AI in business processes and in a corporate culture,” underlines Cyrille Chausson.
Despite these obstacles, French companies plan to increase their investments in AI in the coming months, particularly via the cloud.
What solutions to accelerate your AI projects?
On the revenue side to make up for this delay, IDC confirms the leads from other reports (BCG, Datasulting, Crayon, etc.)
The ingredients recommended by Cyrille Chausson are, jumbled together:
to define a clear strategy with measurable objectives (with KPIs to optimize ROI),
choose between “making” and “making” (in English “Make or Buy”),
adopt an agile and flexible operational model (to encourage experimentation),
have the support of top management,
consider software development and data and AI development in an integrated manner,
invest in training (with internal programs) and recruitment (to equip yourself with the necessary AI skills), create a solid governance framework (which involves putting in place processes, controls and an organization dedicated to compliance , security, privacy protection and ethics), collaborate with trusted partners (notably cloud providers), and finally anticipate the evolution of regulations (through monitoring).
NB Cyrille Chausson was one of the founders of MagIT, of which he was editor-in-chief for several years. He no longer exercises any function in the title today.
