Legal
Disclaimer
Last updated: July 8, 2026
1. Not legal advice
Ravna is an information and research tool. Nothing on this site or in its exports constitutes legal advice, a compliance determination, or a formal regulatory assessment for your organisation. It does not create an advisor–client relationship. Before making compliance decisions you should obtain advice from qualified legal and compliance professionals who can consider your specific deployment, jurisdiction, and vendor configuration.
2. Independent assessment, not statement of fact
Ravna’s risk classifications, category assignments, and obligation notes are our independent interpretation of the EU AI Act and related frameworks, applied to publicly available vendor documentation. They describe how a feature could be classified in a given deployment context. They are reasoned assessments, not assertions of fact, and they are not determinations that any vendor, product, or customer is non-compliant, unlawful, or in breach of any obligation. Reasonable analysts and regulators may reach different conclusions.
3. Point in time; your configuration is authoritative
Findings reflect vendor documentation as it appeared at the assessment date shown on each finding. Vendor features, default states, and documentation change over time. Ravna cannot see inside your tenants: what is actually enabled, how it is configured, and how its output is used are known only to you. Your own environment and records are the authoritative source. Treat each finding as a hypothesis to confirm with your administrators, not as a statement about your live configuration.
4. Vendor names and trademarks
Vendor and product names, logos, and marks referenced on Ravna are the property of their respective owners and are used for identification, reference, and commentary only. Ravna is independent and is not affiliated with, endorsed by, sponsored by, or partnered with any vendor it covers. No vendor funds Ravna or has editorial control over its findings. Reference to a vendor does not imply that the vendor agrees with, or has reviewed, any classification.
5. High-risk and prohibited-practice findings
Some findings identify features that may fall within the EU AI Act’s high-risk categories (Annex III) or, in narrow cases, its prohibited practices (Article 5). Whether such a classification applies depends heavily on how and where the feature is deployed and on the exceptions and conditions in the Act. A high-risk or prohibited-practice label describes a possible classification in a described context; it is not a statement that a vendor’s product is unlawful in all uses, nor that any particular customer is using it unlawfully. These findings in particular should be reviewed with counsel before any reliance.
6. The regulation is evolving
The EU AI Act is recent, and its secondary guidance, harmonised standards, national implementation, and enforcement practice are still developing. Our classifications reflect our reading at the date shown and may change as the legal landscape develops. We update findings on the cadence described in our methodology, but we do not guarantee that any finding reflects the most current legal position.
7. No warranty; limitation of liability
Ravna is provided “as is,” without warranties of any kind, express or implied, including as to accuracy, completeness, or fitness for a particular purpose. To the maximum extent permitted by law, we are not liable for any loss arising from use of or reliance on the service, including regulatory fines, compliance failures, or business losses. This summary is subject to the fuller terms in our Terms of Service.
8. Corrections
If you are a vendor or researcher and believe a finding is inaccurate or addressable by public documentation, we want to know. We review every submission against public sources and correct the record where warranted, with the change logged publicly. Use the corrections form; our review target is five business days.