Inventory every AI system, classify it under the AI Act, run impact assessments and keep the evidence an auditor asks for, in one workspace. Built in Cologne for teams that have to show their work.
Not sure what applies to you? Check your AI Act obligations in two minutes →
AI system register
| System | AI Act classification | Evidence |
|---|---|---|
CV screening model HR · deployer High-risk · Annex III | High-risk · Annex III | FRIA in review |
Customer support chatbot Service · provider Transparency · Art. 50 | Transparency · Art. 50 | Disclosure evidenced |
Demand forecasting Supply chain Minimal risk | Minimal risk | Logged |
Every entry links to its risk assessment, controls and audit trail.
How AIMS works
The AI Act and ISO/IEC 42001 ask the same underlying question: do you know what AI you run, what risk it carries, and can you prove you control it? AIMS turns that into five steps.
Build one inventory of every AI system and model you develop, buy or embed, with owners and intended purpose.
ISO 42001 cl. 4.3 · AI Act Art. 49
Determine your role and each system's risk class: prohibited, high-risk, transparency or minimal.
AI Act Art. 5, 6, 50
Run AI risk and impact assessments, including fundamental rights impact assessments, with the right stakeholders.
ISO 42001 cl. 6.1, A.5 · AI Act Art. 27
Map Annex A controls and AI Act requirements to owners, policies and evidence, and track the gaps.
ISO 42001 Annex A · AI Act Art. 9–15
Run internal audits and management reviews, and export the documentation an assessor will ask for.
ISO 42001 cl. 9.2, 9.3
Two products, one record
AIMS is the system of record for AI governance. Arelis Hub is the gateway your teams use to build with models, and it reports into AIMS.
AI management system
For compliance, risk and data protection teams running an AI governance programme against ISO/IEC 42001 and the EU AI Act.
AI integration platform
For engineering teams that need governed access to models from multiple providers, without building the plumbing.
EU AI Act timeline
As amended by the Digital Omnibus (Regulation (EU) 2026/1744). Several obligations are already in force.
Fines for prohibited practices reach €35 million or 7% of worldwide annual turnover.
Unacceptable-risk AI is banned, including social scoring, manipulative techniques and untargeted facial-image scraping.
Providers of general-purpose AI models must meet documentation, copyright and transparency duties. Fine-tuning can make you a provider.
People must be told when they interact with AI, and synthetic audio, image, video and text must be marked. Applies regardless of risk class.
Risk management, data governance, technical documentation, human oversight and conformity assessment for stand-alone high-risk AI.
High-risk obligations for AI embedded in products covered by EU harmonisation legislation.
Free resources
Practical tools from the team behind the ISO 42001 Implementation Playbook.
Answer a few questions and see which obligations likely apply to a system, with dates and article references.
Classify a system30 editable templates for your AI management system: policy, risk register, Statement of Applicability and more.
Get the templatesA requirement-by-requirement checklist to scope your programme and track progress.
Open the checklist
Who builds Arelis
Arelis AI was founded by Ramón Marrero, author of the ISO 42001 Implementation Playbook for AI Leaders, which walks through every clause of the standard and all 38 Annex A controls.
The team behind AIMS has spent more than a decade building data and AI platforms in production. We built the tool we wanted when we had to evidence AI governance ourselves.
Insights
7 min read
The Digital Omnibus moved the high-risk deadline to December 2027, and a lot of teams heard 'the AI Act is delayed'. Three sets of obligations did not move. Here is what actually applies in August, and what it means for your roadmap.
Ramón Marrero
8 min read
Within a single week, Anthropic and Google released their most powerful AI models yet. Here's what enterprise leaders need to know about Claude Opus 4.5 and Gemini 3—and how to govern these increasingly capable systems responsibly.
Ramón Marrero
10 min read
The EU has pushed back full implementation of high-risk AI rules to December 2027. What this means for businesses, compliance strategies, and the future of AI regulation in Europe.
Ramón Marrero
FAQ
A 30-minute walkthrough of AIMS with your use cases, the AI Act deadlines that apply to you and what an ISO 42001 audit will ask for.
