L'EU AI Act è in vigore. Le scadenze sono attive. hegyon.ai aiuta le PMI svizzere a restare allineate, con guardrail integrati, logging pronto per l'audit e supporto all'AI literacy, in modo che la conformità EU AI Act diventi una funzione della piattaforma che già utilizza.
The EU AI Act enforces in stages. The first deadlines have already passed. High-risk obligations land in August 2026.
Article 4 AI literacy requirements and Article 5 bans on unacceptable-risk systems (social scoring, manipulative techniques) are already enforceable.
Transparency obligations and systemic risk assessments for foundation models and GPAI providers take effect.
The bulk of Annex III high-risk requirements, Article 50 transparency rules, and deployer obligations become enforceable. Fines up to 35M EUR or 7% of global turnover.
High-risk rules extend to AI embedded in regulated products, medical devices, automotive, aviation, and other EU product safety sectors.
La maggior parte delle PMI svizzere sottovaluta ciò che l'EU AI Act richiede loro. Ecco perché.
The AI Act introduces new concepts, risk classification, provider vs deployer roles, Annex III categories, that do not map to existing IT governance frameworks.
Articles 11, 12, and 17 demand technical documentation, automatic logging, and quality management systems that most AI deployments were never built to produce.
Employees use AI tools without IT oversight. Every unmonitored interaction is a compliance liability, and most organisations cannot even inventory their AI systems.
Traditional IT controls were designed for deterministic systems. AI is probabilistic. Bolting compliance onto each system individually does not scale.
hegyon.ai integra le protezioni direttamente, tra le Sue persone, i Suoi dati e i Suoi modelli IA, in modo che traccia d'audit e applicazione delle policy arrivino con la piattaforma che già utilizza, senza diventare un progetto a parte.
Data protection rules, content restrictions, and compliance guardrails enforced automatically at every AI interaction. No manual oversight gaps.
Art. 9 · Art. 10 · Art. 26Every query, response, model invocation, and access event logged with full context. Exportable, auditable, retention-managed.
Art. 12 · Art. 16 · Art. 72Continuous oversight of all AI interactions across your organisation. Policy violations flagged instantly. Incident triggers built in.
Art. 14 · Art. 72 · Art. 73Ospitato nei data center svizzeri di Hostpoint, certificati ISO 27001. I dati sensibili vengono mascherati dal nostro Guard prima che un modello IA esterno li veda. Opzione on-premise disponibile per un isolamento più stretto.
Art. 15 · GDPR · Data SovereigntyFrom initial assessment to continuous monitoring, clear deliverables mapped to specific EU AI Act obligations.
Inventory every AI system in your organisation, determine provider/deployer roles, and classify risk levels according to Annex III.
Configuri il Sensitive Data Guard e le regole di conformità con data protection guard, controllo d'accesso per ruolo e policy di routing dei modelli su misura per la Sua azienda.
Establish logging infrastructure, export formats, and evidence collection pipelines that satisfy Art. 11/12 and prepare you for regulatory scrutiny.
Continuous monitoring with policy-based incident triggers, reporting workflows, and post-market surveillance aligned to Art. 72/73.
A structured, four-phase approach that moves you from regulatory uncertainty to operational readiness.
Inventory AI systems, classify risk levels, determine provider/deployer roles across your organisation.
Set up governance policies, data protection guards, access controls, and model routing rules.
Distribuire il Sensitive Data Guard e la configurazione di conformità, collegare le sorgenti dati, attivare le tracce d'audit e il monitoraggio.
Continuous oversight, incident detection, reporting readiness, and iterative policy refinement.
Direct alignment between platform capabilities and regulatory requirements.
If your organisation uses AI in any of these domains, Annex III high-risk requirements apply by August 2026.
Credit scoring, insurance underwriting, fraud detection, and algorithmic trading systems.
Diagnostic AI, clinical decision support, patient triage, and medical device integration.
CV screening, candidate ranking, performance evaluation, and workforce analytics.
Benefits administration, eligibility assessment, law enforcement support, and border management.
Book a compliance assessment with our team. We will map your AI systems to EU AI Act obligations and show you the fastest path to operational readiness.