Transparency on the use of AI

3.0 version — in effect from 25 September 2026

Our role

CE4U S.r.l. (RAVENG) uses artificial intelligence in two roles specified by the AI Act:

Provider (Article 3(3)) of AI systems that we develop and offer under our brands, built on third-party general-purpose AI models, for example OpenAI, Anthropic and Google;
Deployer (Article 3(4)) when we use AI tools in professional services and internal processes.

None of our systems involve practices prohibited by Article 5 or, for their stated intended purpose, high-risk systems under Article 6. Our systems do not perform facial recognition, biometric identification, emotion recognition, biometric categorisation or automated decisions about people.

AI in professional services (Article 13, Law 132/2025)

In our consultancy services (machinery safety and CE marking, risk assessment, Industry 4.0/Transition 5.0 technical assessments, privacy, cybersecurity and training), we may use AI systems only as supporting tools, for example to draft or summarise text, find regulatory references or check document completeness. Human intellectual work remains predominant. Each document is verified and signed by the responsible professional.

When an engagement is agreed, the client receives clear, comprehensive written information about the systems used and safeguards adopted, and may request that particular activities be performed without AI. Confidential documents and client data are processed only with professional tools that contractually exclude model training, with Article 28 GDPR agreements and safeguards for transfers outside the EU.

AI on this site

Virtual assistant: provides predefined answers selected using keywords. It is not an AI system and does not transmit the text you type to anyone.
Insights articles: some texts are drafted with generative AI assistance. Each article is checked and published under the editorial responsibility of engineer Carlo Tenca (Article 50(4), AI Act), as stated at the end of the article.
Images: the site does not publish images or videos generated or manipulated with AI that realistically depict real people, installations, or places. If we do in the future, they will be visibly labeled as "generated with AI".

AI systems provided by CE4U

System	Intended purpose	Excluded uses	Human supervision
PerizIA 4.0 (perizia40.com)	Drafting and document support for professionals preparing Industry 4.0 / Transition 5.0 technical assessments	Use by or on behalf of judicial authorities; issuing an attestation without a complete review	The expert verifies, integrates and signs; the output is a draft
DVR App / SafetyOS	Support for employers, RSPPs and consultants preparing risk-assessment documents (DVR) and safety and machinery documentation	Assessment or monitoring of individual workers; replacement of the employer's risk assessment	AI proposals are drafts to be verified and approved by the user
Iridai (in development)	Technical inspection assistant for machines and plants	Observation and analysis of people; identification; evaluation of workers	The user decides what to record and validates the transcripts

AI interaction (art. 50, par. 1). In the conversational interfaces of our products (chat and copilot) the user is informed that they are interacting with an AI system and that the responses must be verified.

Marking generated content (Article 50(2)). We are introducing machine-readable metadata in exported DOCX and PDF documents to identify AI-generated or AI-assisted content, following the European Commission's Code of Practice. The transitional deadline for systems already placed on the market before 2 August 2026 is 2 December 2026.

Instructions, known limitations and model notices for professionals using our products are available in the respective documentation and on request.

Internal governance

Corporate AI policy with pre-verification of every new system (prohibited practices, high risk, transparency, privacy, commercial claims).
AI system register with models, suppliers, intended use and obligations, re-examined at least once a year.
AI literacy program for staff and collaborators (art. 4 AI Act).
Data protection impact assessment (art. 35 GDPR) for high risk AI processing.

Reports and contacts

For questions, complaints, malfunctions or inappropriate content generated by our AI systems, write to privacy@raveng.it with AI in the subject. We respond within 30 days. The national AI market-surveillance authority is Italy's National Cybersecurity Agency (ACN; Article 20, Law 132/2025). The Italian Data Protection Authority is responsible for data-protection matters.