EU AI Act compliance for companies that use AI.
Find the AI tools your company already uses, classify each one under the EU AI Act, track the duties that apply with owners and dates, and keep the evidence in one place, ready to hand to a customer, auditor or regulator.
Request early access.
AI Governance is available soon. Leave your work email and we will contact you when early access opens. We use it for nothing else.
Your chatbot is already in scope.
Most AI governance software is built for companies that train their own models. Most companies do not. They buy software, and a growing share of that software quietly includes AI: the support chatbot, the CV screening tool, the writing assistant, the meeting recorder. Under the EU AI Act those companies are deployers, and deployers have duties of their own.
Some of those duties apply today. The Article 50 transparency duties have applied since 2 August 2026 and are not tied to any risk tier: if you run a chatbot or publish AI-generated content, you owe disclosure now. The Digital Omnibus moved the high-risk regime to December 2027. It did not move the rest.
AI Governance is built for that position: the system of record that shows what AI you use, what the Act asks of each system, and what you have done about it.
What applies, and when.
The dates that matter for a deployer, after Regulation (EU) 2026/1744. AI Governance keeps duties that are enforceable now apart from those that are not, so a deadline never reads as a breach.
-
Prohibited practices and AI literacyArticle 5 bans and the Article 4 AI literacy duty.Enforceable now
-
Transparency dutiesArticle 50: chatbot disclosure, labelling of deepfakes and AI-generated content, emotion recognition notices.Enforceable now
-
Annex III high-risk systemsDeployer duties under Article 26 and, where required, a fundamental rights impact assessment under Article 27.Not yet enforceable
-
High-risk AI in regulated productsArticle 6(1) and Annex I systems embedded in products such as machinery or medical devices.Not yet enforceable
What AI Governance is designed to do.
From the first list of tools to the evidence you hand over.
Find shadow AI
Connect Okta, Microsoft 365 or Google Workspace, or paste a list of apps, and see which tools your staff sign in to carry AI you need to account for.
Keep one AI register
Every system you use, buy or embed, with its owner, purpose and source. Ordered worst first, so nothing serious hides in an alphabetical list.
Classify with reasons
A guided questionnaire that follows the Act in its own order. Each step records its article, and a named person confirms the result on a date.
Track what you owe
Classifications become duties with owners and dates, split into enforceable now and not yet enforceable.
Write the notices
Article 50 disclosure text for the systems that actually need it, and AI literacy training records for Article 4.
Record suppliers and incidents
What each supplier gave you and claims, and what went wrong and who you told, in the order the Act sets out.
Map to other frameworks
Crosswalks to ISO/IEC 42001, the NIST AI RMF and the Colorado AI Act, with the gaps listed first.
Export the evidence binder
Your register, classifications, duties, notices and records in one timestamped export, with the approval behind each decision.
Keep history
Classifications are versioned, never overwritten, and every change is logged with who made it and when.
What it deliberately will not do.
In compliance, a confident wrong answer does more harm than no answer. These omissions are the design.
No compliance score
No percentages and no "73% aligned". Counts, dates, owners and evidence only.
No classification from a vendor name
Risk depends on how you use a tool. Discovery asks you a question; it never answers it for you.
No merged deadlines
A duty that is enforceable now is never added to one that starts in 2027.
No legal advice
It organizes your decisions and evidence. Your counsel still signs off on the important ones.
Built for the people who own AI compliance.
AI Governance is designed for companies of roughly 50 to 750 people that use AI in the EU or sell into it: HR and staffing, fintech and lending, insurance, healthcare administration, education, e-commerce and agencies that produce AI content for clients.
It is for the data protection officer or compliance manager who suddenly owns the AI Act, the counsel who needs defensible records, and the engineering lead who has no idea which SaaS tools embed AI.
Need help before it launches? Our compliance and privacy team works on AI inventories, risk classification and policy today.
AI Governance questions.
Does the EU AI Act apply to companies that only use AI?
Yes. The Act separates providers, who build AI systems, from deployers, who use them. Deployers have their own duties. If you run a customer chatbot or publish AI-generated content, the Article 50 transparency duties apply to you whatever the risk tier, and Article 4 AI literacy applies to every organization that uses AI.
Was the EU AI Act delayed?
Partly. Regulation (EU) 2026/1744, the Digital Omnibus on AI, moved the Annex III high-risk regime to 2 December 2027 and product-embedded high-risk systems to 2 August 2028. It did not delay the rest: prohibited practices and AI literacy have applied since February 2025, and Article 50 transparency since 2 August 2026.
What is shadow AI, and how does AI Governance find it?
Shadow AI is AI that staff use without it being on any register: a writing assistant, a meeting recorder, AI inside a SaaS tool you already pay for. AI Governance reads the apps people actually sign in to from Okta, Microsoft 365 or Google Workspace, or from a list you paste in, and flags the ones that carry AI you need to account for.
Does AI Governance decide whether a system is high risk?
No. Whether a system is high risk depends on how you use it, not on what the vendor sells. The classification wizard walks you through the Act's own questions and records the reasoning, but a named person in your organization confirms every classification.
Does it give us a compliance score?
No, on purpose. A percentage such as 73% aligned is a compliance claim nobody can stand behind. AI Governance shows counts, dates, owners and evidence, and keeps duties that are enforceable now apart from duties that are not enforceable yet.
Is AI Governance legal advice?
No. It is software that organizes your AI inventory, classifications, duties and evidence. It does not replace qualified legal advice, and we recommend your counsel reviews important classifications.
How do I get early access?
AI Governance is available soon. Leave your work email on this page and we will contact you when early access opens.
Get AI Governance early.
Request early access and we will contact you when places open. Early users help shape what the product covers first.