|
|
The EU AI Act has been taking effect in stages since 2025; since August 2026 the transparency obligations for AI-generated content (Article 50) apply, among others. Many companies already use AI but do not know which obligations follow from it. FHC+P brings both together: we make your AI use productive AND compliant, pragmatic, documented, and without slowing your teams down.
ChatGPT in sales, AI images in marketing, Copilot in development: in most companies AI is already in use, often as uncontrolled shadow AI. The EU AI Act turns this into a leadership task: AI literacy for employees (Article 4), transparency and labelling obligations for AI content and chatbots (Article 50), prohibited practices, and, depending on the use case, high-risk requirements. Those who wait risk fines and sprawl. Those who over-regulate lose the productivity advantage. The right path lies in between, and with clear guardrails it is very achievable.
Where is AI used in your organisation, officially and unofficially? We capture all use cases and map them to the AI Act risk classes: prohibited, high-risk, transparency-bound, or free. The result is your prioritised action list.
Labelling AI content, machine-readable marking of AI images, transparency for chatbots: we translate the obligations into concrete processes and tools, including approval and veto workflows through which reviewed content carries editorial responsibility.
Since February 2025 the AI Act requires sufficient AI literacy from everyone who uses AI. We train your teams close to their real day-to-day work: what they may do, what they may not, and how to spot risks, without losing the appetite for innovation.
An AI policy that is read and lived: permitted tools, data classes, approval processes, responsibilities. Lean documentation, interlocked with GDPR, and clear escalation paths instead of a paper tiger.
Which AI tools meet your requirements for data protection, EU processing, and traceability? We select tools, configure guardrails, and replace shadow AI with better, permitted alternatives.
For us compliance is not a brake but part of the AI transformation: the same analysis that identifies obligations also finds the use cases with the greatest benefit. You get both in one approach.
Our own content publishing platform Personativ publishes AI-supported content in an Article-50-safe way: human-reviewed posts carry editorial responsibility and need no AI label, unreviewed ones are labelled automatically, AI images are marked machine-readably, and the approval log provides the proof. We bring this implementation experience, from legal requirement to working workflow, into your organisation. Note: we deliver the technical and organisational implementation; legal advice on individual cases is provided by your law firm, with whom we are happy to collaborate.
What applies since February 2025, August 2025, and August 2026, and what comes by August 2027, with the to-dos per stage.
To the overview → Self-checkIs your AI content labelling-safe? The self-check shows in 10 questions where you stand and what is still missing.
To the self-check →Start with an AI inventory: within a few days you will know where you stand, what the AI Act requires of you, and which opportunities remain untapped.
Have further questions? Feel free to contact us directly via our contact form or by e-mail.
To practically every company that uses or provides AI systems, regardless of size. The obligations depend on the risk of the use case: from simple transparency duties for chatbots and AI content to extensive requirements for high-risk systems. Even pure users must ensure AI literacy.
The AI Act takes effect in stages: since February 2025 the prohibitions and the AI literacy duty (Article 4) apply, since August 2025 the rules for general-purpose AI models, and since August 2026 the transparency and labelling obligations (Article 50), among others. Further stages follow. Starting today means working through the obligations in an orderly way instead of under pressure.
Transparency: people must be able to recognise when they interact with an AI, and AI-generated or manipulated content must be recognisable or marked machine-readably. Exemptions apply to editorially reviewed content under human responsibility, which is exactly why clean approval workflows with proof are needed.
No. FHC+P delivers the technical and organisational implementation: inventory, risk classification, processes, tools, training, and documentation. The legal review of individual cases is done by your legal department or law firm, with whom we collaborate closely and gladly.
Because we have implemented the requirements productively ourselves: our own platform Personativ publishes AI content in an Article-50-safe way, with an approval and veto principle, automatic labelling of unreviewed posts, and machine-readable marking of AI images. This hands-on experience considerably shortens your path from requirement to working solution.