Transparency & governance
AI Act compliance statement
Version 1.0 · 10 October 2026
Purpose and status
CertifyAI supports documentation and governance under Regulation (EU) 2024/1689. This statement describes the app’s current safeguards and limitations; it is not a declaration of conformity, CE marking, legal opinion or proof of certification. No independent AI Act conformity assessment of this app is claimed. The operator’s legal identity must be confirmed before production use.
You are interacting with AI
The Help AI helper uses a generative OpenAI language model through Lovable’s AI gateway. Its replies are AI-generated, may be inaccurate, incomplete or out of date, and must be reviewed. It provides product guidance, not binding legal advice or official regulatory interpretation. This disclosure supports the interaction-transparency objective of Article 50; it does not alone establish compliance with all applicable obligations.
Rules-based classification, not certification
The Classification Engine follows a rules-based question tree covering Article 5, Article 6, Annex I, Annex III and Article 6(3). It is distinct from the generative AI helper. A memo reflects the user’s answers, rationale and version. Readiness scores measure documentation progress, not legal conformity. Evidence lifecycle states and ledger references establish workflow traceability, not approval by an authority or Notified Body. Simulated anchoring is not proof of a live ledger transaction.
Human oversight and access controls
Users remain responsible for verifying source law, dates, evidence and classification outcomes. Qualified compliance and legal professionals should review any consequential determination. Authoring, signing, validation and approval are role-controlled; the AI helper does not change compliance records, billing or approvals. Its server receives the conversation history needed for guidance, not an automatic copy of workspace compliance evidence.
Intended use and risk assessment
The intended purpose is internal compliance workflow assistance, not decisions about employment, credit, healthcare, biometrics, policing or access to essential services. We do not claim that the app is categorically exempt from high-risk obligations. Provider/deployer roles and classification must be assessed against actual functionality, deployment context and intended use. Using the app to support a high-risk system does not itself certify that system. Material changes or deployment for a regulated use require a fresh assessment.
AI literacy and responsible use
Organisations should provide role-appropriate AI literacy measures under Article 4, review AI outputs, prohibit submission of unnecessary sensitive data, and maintain oversight and incident procedures. Do not rely on the helper for emergency, medical or autonomous legal decisions. Report harmful or misleading guidance with the relevant conversation context, without including confidential data.
Data protection and remaining obligations
See the privacy notice for conversation processing, providers and data rights. Before production use, the operator must confirm contractual responsibilities, provider documentation, retention and transfer safeguards and the applicability of any further AI Act requirements. Customers must assess their own obligations, including risk management, technical documentation, quality management and conformity assessment where applicable. Displayed deadlines should be checked against current official law.
Questions and concerns
Contact andrew@ripponconsulting.com for AI governance questions or to report a concern.