In 2024, Insee measured that 10% of French companies use at least one artificial intelligence technology, compared to a significantly lower proportion a year earlier. This figure, modest at first glance, reflects a shift: the technologies that have fueled strategic discussions for several years are now entering operational budgets.
The landscape of technological trends in 2024 is not just a list of spectacular innovations. It reflects concrete trade-offs between regulation, cybersecurity, cloud costs, and the actual adoption of artificial intelligence by companies.
When the technological trend becomes a budget item
Listing technological trends each year often produces the same effect: a catalog of promises that look alike from year to year. Artificial intelligence, 5G, quantum, blockchain – these terms were already present in the reports of 2022 and 2023.
The difference in 2024 lies less in the technologies themselves than in their transition from exploratory to budgetary status. A company that was testing a machine learning prototype in 2023 must now finance its industrialization, recruit specialized profiles, and adapt its information systems. This shift changes the very nature of what we call a “trend.”
Several sectors in France illustrate this shift. Financial services, insurance, and healthcare are among the first to integrate AI software into their business processes, no longer in the lab but in production. Finding tech articles on Web Adresses allows you to follow this transition sector by sector, beyond generic announcements.
Artificial intelligence in business: beyond text generation
Generative artificial intelligence has monopolized media attention since late 2022. In 2024, the use of AI in business far exceeds content generation. Applications are diversifying towards the analysis of structured data, automation of administrative tasks, and decision support in sectors such as insurance or logistics.
Machine learning, in particular, is gaining ground in anomaly detection systems. An industrial company can now monitor its production lines with algorithms trained on its own historical data, without relying on a generic model.
What companies underestimate in deploying AI
Adopting artificial intelligence software is not limited to purchasing a license. The quality of training data determines the reliability of the results. A poorly structured or incomplete database produces unusable predictions, regardless of the model used.
The other point of friction concerns governance. Who validates the decisions made by an algorithm? Who bears responsibility in case of an error? These questions, long theoretical, become operational as soon as an AI system intervenes in a critical process.
- Data preparation often represents the longest phase of an AI project, well before the training of the model itself
- Companies that succeed in their deployment designate an internal referent responsible for supervising algorithms in production
- The actual cost of an AI project includes the ongoing maintenance of the model, which must be retrained as data evolves
Cybersecurity and AI: a two-way race
Cybersecurity in 2024 is characterized by a technical paradox. The same artificial intelligence technologies that strengthen companies’ defenses are also used to sophisticate attacks. Cybercriminals use AI to automate phishing and bypass detection systems.
This escalation forces companies to rethink their approach. Simply installing a firewall or antivirus is no longer sufficient. Defense systems must integrate machine learning capable of identifying abnormal behaviors in real-time, on data volumes that human analysis can no longer process.
Regulation and compliance as accelerators
The European NIS 2 directive, applicable in 2024, imposes strengthened obligations for incident notification and risk management on a broader range of companies. This regulatory framework pushes sectors that were previously less concerned (industrial SMEs, digital service providers) to invest in security technologies that they would have otherwise postponed.
Cloud FinOps: managing costs rather than migrating
Migration to the cloud is no longer a trend in itself. The majority of large French companies have already transferred a significant portion of their applications and data to cloud infrastructures. The question in 2024 is no longer “should we migrate?” but “how to control the bill once the migration is done?”.
This is where Cloud FinOps comes into play, a discipline that combines financial engineering and cloud infrastructure management. The goal: to give technical and financial teams shared visibility on the actual consumption of resources, to eliminate unnecessary expenses.
- Rightsizing involves adjusting the power allocated to actual needs, rather than keeping costly safety margins
- Automatically shutting down test environments outside of working hours reduces the bill without affecting production
- FinOps dashboards allow each cloud expense item to be attributed to a team or project, making budgetary trade-offs more precise
CSRD and technologies: when ESG reporting dictates software choices
The CSRD (Corporate Sustainability Reporting Directive) has pushed French companies to structure their collection of environmental and social data starting in 2024. This regulatory constraint has a direct effect on technological choices: information systems must now integrate extra-financial reporting modules.
ESG compliance transforms management software into strategic steering tools. Companies no longer choose an ERP solely based on its accounting or logistical functions, but also on its ability to produce the indicators required by the CSRD.
This link between technological innovation and sustainability is likely one of the highlights of 2024. Technologies are no longer evaluated solely on their performance or cost, but also on their ability to meet transparency obligations that did not exist two years ago. This year’s technological news can be read as much in regulatory texts as in product announcements.