The year 2024 has reshuffled the technological priorities of companies. The pilot projects for generative AI launched in 2023 are now facing an integration wall, while new layers of automation and regulatory frameworks are changing the game for technical teams.
Agentic AI: Automation that Goes Beyond Content Generation
Text and image generation is no longer the frontier of artificial intelligence in business. We are witnessing a shift towards agentic AI, designed to execute end-to-end business workflows. An agent does not just produce a response: it chains actions, queries databases, triggers processes, and adjusts its behavior based on intermediate results.
This technical distinction has direct consequences on software architecture. An agentic pipeline requires reliable orchestration between calls to language models, internal APIs, and data management systems. Teams that deployed a chatbot in a few weeks are discovering that transitioning to an autonomous agent requires a redesign of connectors, access rights, and oversight mechanisms.
To keep up with the tech news on Mister Free Free, this agentic shift represents the topic to watch closely, as it conditions the real value that organizations will derive from their investments in artificial intelligence.
Generative AI Pilots and Return on Investment: The Real Blockage
The adoption of generative AI has continued to progress, moving from a confined use in a business function to a broader dissemination within organizations. The problem is no longer experimentation. The majority of AI pilots are not yet transforming the P&L, and the reasons are rarely related to the models themselves.
The identified barriers are structural:
- Fragmented data across multiple systems, with incompatible formats and insufficient governance to reliably feed a model
- Business processes designed before the AI era, where adding a generative model only accelerates an already suboptimal workflow
- A skills gap between data scientists and operational teams, which slows down production deployment
We recommend addressing the data problem upstream, before multiplying use cases. A high-performing model fed by poorly structured data produces unusable results in a business context.
Cyber Resilience Act: Cybersecurity Becomes a Product Obligation
The Cyber Resilience Act, adopted at the end of 2024, imposes security requirements on most digital products marketed in the European market. The obligations will be phased in until 2026-2027, leaving a tight deadline for software publishers and manufacturers of connected devices.
This text changes the usual logic. Until now, cybersecurity was a matter of architectural choice or internal policy. With this regulation, security by design becomes a market entry condition. Products will need to integrate update mechanisms, vulnerability reporting, and technical documentation from their design stage.
Concrete Impact on Development Cycles
Development teams will need to integrate compliance audits into their sprints. Companies distributing connected devices or consumer applications are the first affected, but the scope of the regulation also covers embedded software components.
For organizations that have not yet formalized their vulnerability management process, the timeline requires action now. Waiting until the last months before the regulation takes effect risks a commercial blockage.
Quantum Computing and Post-Quantum Cryptography: A Discreet but Urgent Task
The term “cyberquantum” is circulating in forward-looking reports, but the concrete issue for 2024 lies on the defensive side. Post-quantum cryptography prepares for the transition before quantum computers break current standards. NIST has finalized its first standards for resistant algorithms, and major cloud providers are starting to offer compatible implementations.
The main risk concerns data encrypted today and potentially decryptable tomorrow. Organizations handling sensitive long-lived data (health, defense, intellectual property) must plan the migration of their cryptographic protocols without waiting for the availability of a large-scale functional quantum computer.
Technical Priorities for CIOs
- Inventory the cryptographic algorithms used across the entire information system
- Identify data flows whose confidentiality must be guaranteed for more than ten years
- Test standardized post-quantum algorithms in the most exposed environments before a global deployment
This task does not produce visible short-term results, but it conditions the digital resilience of the coming years. Companies that postpone this transition expose themselves to costly catch-up when the quantum threat becomes operational.
The technological trends of 2024 share a common thread: the shift from demonstration to industrialization. Agentic AI replaces showcase chatbots, the Cyber Resilience Act transforms security into a regulatory prerequisite, and post-quantum cryptography requires anticipating a still theoretical threat. Technical teams addressing these three issues in parallel are positioning themselves on the foundations of the architectures of the next decade.
