What indicators allow us to measure the technological changes that have truly impacted the web and businesses in 2024? Between the massive deployment of generative artificial intelligence, the tightening of European regulations, and the evolution of development tools, the available data paints a picture where the speed of adoption matters less than the ability to comply with the new rules of the game.
Mandatory marking of AI content: what the AI Act changes for web actors
Most tech assessments for 2024 focus on the performance of generative models. The legal framework governing them deserves equal attention, as it directly conditions how these technologies can be deployed on a site or platform.
The Article 50 of the European AI Act has required since August 2026 that providers of large generative models technically mark all content produced or manipulated by AI and intended for the EU market. Watermarking, signed metadata, digital tattooing: the technical choice remains open, but the obligation is firm.
The compliance deadline runs until December 2, 2026, for systems already on the market. The penalties can reach 15 million euros or 3% of global revenue. For companies using AI-generated visuals on their e-commerce sites, marketing carousels, or chatbots, the issue is no longer technical but legal.
Meanwhile, the Digital Services Act has placed ChatGPT, Reddit, and Roblox under its strictest regime after these platforms reported exceeding 45 million monthly users in the EU. Enhanced moderation, algorithmic transparency, systemic risk assessment: obligations are piling up for very large platforms that integrate artificial intelligence components.
For those wishing to access the Planet X Tech site and follow these regulatory developments applied to the web, the intersection of the AI Act and DSA is the main point of vigilance during this period.
Generative artificial intelligence and development tools: adoption overview in 2024
The prevailing discourse presents generative AI as a universal accelerator. Feedback from developers themselves paints a more nuanced picture.
| Indicator | Data | Source |
|---|---|---|
| Developers considering coding with AI “more addictive than useful” | 80% | ZDNet / Briefia |
| Platforms subject to the strict DSA regime (examples) | ChatGPT, Reddit, Roblox | Blog du Modérateur |
| Threshold of monthly EU users triggering enhanced DSA regime | 45 million | European Commission |
| Maximum fine under AI Act (marking AI content) | 15 M€ or 3% global revenue | AI Act, art. 50 |
The figure that stands out: 80% of developers find AI-assisted coding more addictive than useful. The perceived productivity gain does not necessarily translate into code quality or bug reduction. The rapid adoption of code assistants (GitHub Copilot, Codeium, Cursor) masks a cognitive dependency issue that technical teams are beginning to document.
Gap between adoption and real mastery
The enthusiasm for generative AI solutions in web development does not translate into uniform mastery. Companies that integrate these tools without appropriate code review processes expose themselves to silent regressions. Production speed increases, but so does the time spent debugging.
In contrast, teams that use AI as a rapid prototyping tool, with systematic human validation, report tangible gains in exploratory phases. The key lies in internal governance, not in the technology itself.
Combined DSA and AI Act: new constraints for platforms and e-commerce sites
The intersection of these two European regulations creates a new zone of constraints for web actors in 2024-2026. Three concrete obligations deserve to be highlighted:
- All content generated or manipulated by AI (product image, automated description, chatbot response) must bear a technical marking readable by verification systems, under penalty of heavy financial sanctions
- Platforms exceeding 45 million monthly users in the EU must publish transparency reports on their algorithmic systems and assess systemic risks (misinformation, impact on minors)
- E-commerce sites using AI-generated visuals or product sheets must adapt their production chain to integrate digital tattooing before the December 2026 deadline
For French companies, simultaneous compliance with the DSA and the AI Act represents a new budget line. Technical marking solutions exist (C2PA, invisible watermarking), but their integration into CMS and e-commerce platforms is still being standardized.
Web technologies 2024: where the real gaps lie
The most publicized trends (WebAssembly, Progressive Web Apps, headless architecture) are progressing, but their adoption remains concentrated among large organizations with dedicated technical teams.
The most significant gap lies between the available technologies and their regulatory compliance. A site deploying a chatbot based on a generative model must now simultaneously manage technical performance, user experience, and the regulatory marking of AI-generated responses.
Cloud and data: a field of increasing tension
Cloud solutions continue to shape web infrastructure, but data localization and digital sovereignty are becoming selection criteria on par with performance. New technologies are no longer deployed in a legal vacuum.
Generative artificial intelligence, sovereign cloud, and the constraints of the DSA form a triangle that each company must navigate according to its exposure to the European market. Regulatory compliance now conditions the timeline for adopting web innovations, not the other way around.
The defining fact of this period remains the shift from an ecosystem where technology preceded regulation to a model where European rules set the pace. Companies that integrate this constraint from the design phase of their web projects will save time compared to those that will have to adapt their tools afterward.
