EU AI Act: demonstrably in order before 2 August 2026
We help operations and compliance teams in food, logistics and greenhouse horticulture implement the EU AI Act in a practical, demonstrable way, without bringing the operation to a standstill. AI advises, you decide.
The bridge between the law and your operation · aligned with HACCP and FSSC
- The EU AI Act regulates AI by risk. Most operational applications mainly require transparency, human oversight and demonstrability, not a ban.
- The AI literacy duty in Article 4 has applied since 2 February 2025; Dutch supervision by the AP and the RDI takes shape from 2 August 2026.
- We make the law practical: inventory, risk classification and audit-ready documentation, aligned with HACCP and FSSC, with the human who decides and signs.
What the EU AI Act means for your operation
The EU AI Act is the first European law that regulates AI, based on risk. The law applies not only to those who build AI, but also to those who use it. In the fresh and food chain that touches a great deal: route optimisation in planning, demand forecasting for purchasing, image recognition for quality control on the line and smart control of cooling and climate. The higher the risk, the heavier the requirements: inventory, classification, documentation, human oversight and monitoring. AI embedded in machines is covered as well.
Do you use ChatGPT, Copilot or similar tools? These are general-purpose AI models (GPAI). Since August 2025, transparency requirements apply to them. We include this use in the inventory, so nothing stays out of sight here either.
The obligations take effect in phases
1 Aug 2024
The EU AI Act entered into force.
2 Feb 2025
Prohibited practices and AI literacy: your team must have basic knowledge of AI.
2 Aug 2025
Obligations for general-purpose AI models (GPAI), such as ChatGPT and Copilot.
2 Aug 2026
Dutch national supervision starts, with the Autoriteit Persoonsgegevens (AP) and the RDI as key supervisors. This is the turning point for most operations.
2 Dec 2027
High-risk AI systems under Annex III must comply, moved from 2 Aug 2026 by the Digital Omnibus.
2 Aug 2028
High-risk AI embedded in regulated products (Annex I) must comply.
These dates follow from the Digital Omnibus amendment to the EU AI Act, finalised on 29 June 2026. High-risk AI under Annex III moves from 2 August 2026 to 2 December 2027; Dutch supervision by the AP and the RDI still starts on 2 August 2026. See the sources at the bottom of this page. Not legal advice.
Where does the risk sit in your operation?
Whether a system is high-risk depends on the application. These are common examples in food, logistics and greenhouse horticulture, with an indication of the likelihood and what you then arrange.
Route optimisation and ETA
Medium to highWhy: Can affect safety and on-time delivery.
You arrange: Risk management and human oversight.
Quality control with image recognition
HighWhy: Co-decides on product safety on the line.
You arrange: Data quality, documentation and sample checks.
Predictive maintenance
MediumWhy: Depends on the impact on machine and food safety.
You arrange: Risk assessment and logging.
Smart climate and cooling control
MediumWhy: Can affect food safety and the cold chain.
You arrange: Monitoring and logging.
AI in recruitment and workforce planning
HighWhy: Affects the rights of employees.
You arrange: Transparency and human oversight.
Indicative assessment, not a legal classification. Have the final assessment legally reviewed.
Your preparation for August 2026
The minimum steps to be demonstrably in order. We do this with you, in 4 to 8 weeks.
- Inventory your AI systems, including AI embedded in machines
- Create a risk classification: high, limited or minimal risk
- Set up a Risk Management System
- Record procedures for human oversight
- Set up logging and documentation
- Train your team in AI literacy (required since February 2025)
- Only need to cover the training obligation? See the AI literacy training, demonstrably arranged in one day.
Your compliance dossier in one place
You work in a secure dashboard. You inventory your AI systems, receive an AI-drafted risk classification and documentation as a draft, review and approve it yourself, and export audit-ready documents to PDF and Word.
- Status per system: compliant, in progress or high risk
- Only approved documents can be exported
- A chronological audit log as evidence for your auditor
What you get
AI inventory
All AI systems in your operation mapped, including AI embedded in machines.
Risk classification
Classification under the EU AI Act, with a sober rationale per system.
Human oversight
A Risk Management Plan and procedures for human oversight, tailored to your operation.
Audit-ready documentation
Documentation that fits your HACCP and FSSC routines.
Implementation roadmap
What needs to happen when, with owners and review moments.
Training
AI literacy and oversight for your team, as the law has required since 2025.
How it works
- 01
Practice Scan or introductory conversation
Together we determine where AI sits in your operation and where the risks are.
- 02
Inventory and classification
Each AI system receives a provisional risk classification with a rationale.
- 03
Drafting the documentation
Drafts for classification, risk management and oversight. You review and sign.
- 04
Training and roadmap
A short training for your team and a concrete implementation roadmap.
- 05
Optional maintenance
Annual updates for audits, changes and new AI systems.
Who it's for
Directors, operations managers, plant managers and quality or compliance managers at food manufacturing, cold-chain logistics and greenhouse horticulture companies with 50 to 500 employees.
Track
4 to 8 weeks
€6,500 to €9,500
Annual maintenance
optional
€2,500 to €4,000
Prices are indicative, excluding VAT.
The result: less risk, faster audits and AI that is actually used on the work floor.
Frequently asked questions about the EU AI Act
What is the most important deadline?
For most companies, 2 August 2026 is the next turning point: that is when Dutch supervision of the EU AI Act starts, carried out by the Autoriteit Persoonsgegevens (AP) and the RDI. Following the Digital Omnibus amendment, finalised on 29 June 2026, high-risk AI systems under Annex III do not need to comply until 2 December 2027, and high-risk AI in regulated products under Annex I until 2 August 2028. The Article 4 AI-literacy duty has applied since 2 February 2025.
Are ChatGPT and Copilot covered?
Yes, as general-purpose AI models (GPAI). Since August 2025, transparency requirements apply to such models. If your team uses generative AI in the operation, we include that in the inventory.
Does my company fall under the EU AI Act?
The EU AI Act applies to organisations that develop or use AI in the EU. If you use AI in your operation, for example route optimisation, demand forecasting or quality control with image recognition, that use is likely covered. AI embedded in machines counts as well.
What is a high-risk AI system?
A system that carries heavier obligations, for example around documentation and human oversight. Whether your system is high-risk depends on the application. We make a first, provisional assessment and clearly flag what needs legal review.
How long does a track take and what does it cost?
A track takes 4 to 8 weeks and costs €6,500 to €9,500, depending on the scope and the number of AI systems. Optional annual maintenance is €2,500 to €4,000 per year.
Do you provide legal advice?
No. We deliver practical templates, documentation and guidance, and act as the bridge between the law and your operation. We advise you to have the drafted documents legally reviewed before they are finalised.
This timeline is based on the following public sources, which you can verify yourself:
Demonstrably in order before 2 August 2026, without bringing the operation to a halt.
Start with the free Practice Scan or book a 30-minute introductory conversation.
We deliver practical templates and guidance, not legal advice. Have the drafted documents legally reviewed.