Blog

News, guides and analysis on AI Act compliance

AI Act Article 50marquage contenu synthétiqueC2PA standard

Anthropic Rolls Out Invisible C2PA Watermarking: What the AI Act Requires from August 2026

Anthropic announces the systematic integration of the C2PA standard into its Claude models, a first for a major generative AI provider. This decision, effective ahead of the 2 August 2026 deadline, anticipates the transparency obligations under the AI Act. The article examines the technical and regulatory implications for providers and deployers, as well as the limitations of this approach in meeting EU requirements.

19 August 20267 min

AI Act recrutementobligations DRH IAsystèmes IA haut risque RH

AI Act and recruitment: what HR directors must know in 2026

Since 2 August 2026, the AI Act has imposed strict obligations on HR directors using AI tools for recruitment. Automated CV screening, candidate scoring, or emotional analysis during interviews: these practices are now regulated or even prohibited. This article details the obligations, prohibitions, and opportunities for HR departments, with a clear table of permitted, prohibited, or conditional use cases.

17 August 20266 min

ISO 42001 AI Actconformité système management IAcorrespondances ISO 42001 AI Act

ISO 42001 and AI Act: Aligning Compliance and Certification

ISO/IEC 42001:2023 and the AI Act form a complementary framework for European organisations. This article examines the correspondences between the two standards, explains why the ISO standard serves as the provisional benchmark in the absence of harmonised standards, and details key differences. A practical guide for compliance officers and auditors seeking to structure their regulatory approach.

13 August 20268 min

AI Act comparaison internationaleréglementation IA UE vs USA vs Chineconformité IA multinationales

AI Regulation: Navigating the EU, UK, China, and US Frameworks

AI is governed by divergent regulations across regions. The EU enforces a binding regulation (AI Act), the UK adopts a sectoral approach, China targets specific sectors, and the US balances innovation with fragmented regulation. This article compares these frameworks to help multinational companies adapt their compliance strategies.

11 August 20267 min

Article 50 AI Acttransparence IA générativeobligations déployeurs IA 2026

Article 50 of the AI Act: Practical Impacts for Businesses with Real-World Examples

Article 50 of the AI Act comes into force on 2 August 2026. Beyond general obligations, this article examines its concrete impacts on business processes, compliance costs, and legal risks for providers and deployers. Featuring sector-specific use cases (media, finance, healthcare, HR) and an analysis of available tools to meet transparency and watermarking requirements.

10 August 202612 min

investissement IA entrepriseconformité AI Act 2026stratégie IA réglementaire

Investing in AI for Business: Balancing Performance and Compliance

The AI Act establishes a strict framework for AI investments, but far from being a constraint, this regulation can become a performance driver. Discover how to structure your AI projects to maximise return on investment while meeting legal obligations, with concrete sector-specific examples and a proven methodology for prioritising your initiatives.

8 August 202610 min

documentation technique AI ActAnnexe IV AI Actconformité systèmes IA haut risque

AI Act Technical Documentation (Annex IV): A Practical Guide to Preparing Your Compliance File

Annex IV of the AI Act requires comprehensive technical documentation for high-risk AI systems. This practical guide details the 9 regulatory pillars, common pitfalls, and a methodology to transform these requirements into verifiable technical deliverables. Learn how to structure your compliance file with auditable technical artefacts and download the free Annex IV Workbook (23 pages, 48 controls) to get started immediately.

5 August 20268 min

AI Act obligations employeursdroits salariés IA travailinterdictions AI Act employeurs

Your employer and AI: what the European law now prohibits

Since 2 August 2026, the AI Act strictly regulates the use of AI in the workplace. Discover what your employer is no longer allowed to do, from prohibited practices to transparency obligations, and how to check if your work tools comply with EU regulations.

4 August 20266 min

Article 50 AI Acttransparence IA générativeobligations déployeurs IA

AI Transparency: What Article 50 Really Requires from Businesses

Article 50 of the AI Act introduces transparency obligations for generative AI systems from November 2026. This article details the concrete requirements for deployers and providers, synthetic content marking methods, and non-compliance risks. Featuring sector-specific examples and an analysis of the latest clarifications from the AI Office.

3 August 20266 min

AI Act PMEchecklist conformité IAobligations AI Act 2026

AI Act Checklist for SMEs: 10 Priority Actions for 2026

The European legislator has considered SMEs in the AI Act, with 38 explicit mentions and support measures such as reduced fees or priority access to regulatory sandboxes. This operational checklist details the 10 priority actions to take in 2026, ranked in chronological order: from inventorying AI tools to preparing the Fundamental Rights Impact Assessment (FRIA). Each step specifies the responsible party, estimated time, and relevant AI Act article.

12 June 20267 min

AI Act pratiques interditesArticle 5 AI Actsanctions AI Act 35 millions

35 million euro fine: 6 AI practices banned from 2025

From 2 February 2025, six AI practices are formally banned in Europe, with penalties of up to 35 million euros or 7% of global turnover. This article details each prohibition under Article 5 of the AI Act, including concrete examples, affected tools, and guidance to verify your organisation's compliance.

8 June 20266 min

AI Act Article 50transparence IA générativeobligations fournisseurs IA

Article 50 of the AI Act: What Actually Applies on 2 August 2026

2 August 2026 marks the entry into force of the transparency obligations under Article 50 of the AI Act. Contrary to common misconceptions, the Omnibus agreement did not postpone all requirements to 2027. Only the machine-readable marking of synthetic content benefits from a four-month extension until 2 December 2026. This article clarifies the four affected scenarios, the roles of providers and deployers, and key steps for compliance.

1 June 202610 min

1 2 3 4