AI Act transparency obligations are in force

Know which AI is in your stack, and which duties land on you.

An independent inventory of the AI features inside enterprise software, mapped to the EU AI Act obligations that fall on you as the deployer rather than on the vendor. Every finding shows its source, and says so plainly where we could not establish it.

105

Vendors covered

255

AI features documented

85

Classified high-risk

The deadline moved. Transparency didn't.

Website transparency check

Five questions. See the Article 50 disclosures your site owes by 2 August 2026.

Check your site

EU AI Act timeline

What actually applies when, after the Digital Omnibus deferred high-risk to 2027.

See the timeline

Disclosure templates

Ready-to-paste AI chatbot, AI-content, and deepfake notices.

Get the copy

The problem

The compliance gap nobody is talking about.

EU AI Act obligations fall on deployers — the organisations that use AI systems in their operations. But most compliance tools focus on the vendor's posture, not on what features you have running by default in your environment.

01

Vendors enable AI features by default

Most enterprise AI features are on when you sign the contract. You may be running high-risk AI systems without having made an active decision to do so.

02

You are the deployer. You are liable.

Under EU AI Act Article 26, deploying organisations carry their own obligations for high-risk AI systems — human oversight, monitoring and record-keeping — regardless of who built them. Article 16 covers what the provider owes; Article 26 is the part that lands on you.

03

Nobody publishes this information

Vendor documentation doesn't clearly state which features are high-risk under the Act, or what the default state is in a fresh tenancy. Ravna does that research independently.


How it works

From heatmap to obligation exposure in three steps.

1

Browse the heatmap

See every AI feature we've classified across your vendors, with default state, risk level, and the specific deployer obligations it triggers.

Open heatmap
2

Mark your stack

Sign in and mark which features are active in your environment. Add deployment context — who's affected, what decisions it influences.

Get started
3

See your obligations

Your portfolio shows adjusted risk based on your context, the real count of EU AI Act obligations you've triggered, and a remediation workflow.

Build your stack

Independent

No vendor funding or editorial access

Versioned

Every finding carries its methodology version

10-day corrections

Disputed findings reviewed within 10 business days

Read the methodology →

August 2, 2026

The deadline has passed. Know your exposure.

AI Act transparency rules apply 2 August 2026; high-risk obligations follow in December 2027. Start with the heatmap, it's free, no account required.

Open heatmapCreate free account