Occupational safety, environmental management, and sustainability are highly regulated fields with extensive requirements for documentation, assessment, communication, and disclosure. A wide range of tasks and stakeholders must be managed internally and externally. To act strategically, measurable targets and KPIs are essential.

Corresponding data collections for EHS and ESG must be well structured and properly maintained. AI in EHS offers enormous potential to speed up data capture, enhance existing data, and unlock its full informational value. Beyond automating processes and saving valuable time for specialists, AI enables entirely new insights and a shift from reactive to proactive approaches – especially in risk management. 

Quentic is a cloud-based, all-in-one solution for EHS and sustainability, and its integrated AI capabilities are purpose‑built for these use cases. For example, you can extract key information from safety data sheets with a simple file upload, analyze inspection videos for safety issues, or accelerate root cause analyses with pre-filled suggestions. The Quentic team places strong emphasis on developing and integrating EHS AI features and will continue to roll out more capabilities and innovations.

Quentic is a modular software platform that you can tailor to your needs. You can add AI-enabled functionality either to the core system features or to individual software modules. Each module covers a specific EHS or sustainability domain – for example, hazardous chemicals management. We are happy to advise you whether you are implementing Quentic for the first time or expanding your existing system.

Quentic’s EHS AI features are integrated directly into the Quentic user interface and require no specialized knowledge. Depending on the use case, you can trigger standardized, automated processes with a single click and receive results instantly. In other situations, you interact with an AI assistant – for example, when navigating legal topics through simple question‑and‑answer dialogue. You can also activate central system settings to receive recurring reports conveniently by email.

Yes. Your data in Quentic is strictly protected, meaning only your authorized user group can access it. Our AI tools are technically embedded in the Quentic system and process and store your data exclusively within the Quentic environment used for application operations and data hosting.

No. Quentic does not use customer data, prompts, or outputs to train, fine-tune, or improve general-purpose AI models. This is a hard requirement in our AI governance – no exceptions. Where we use third-party AI infrastructure, we contractually and technically ensure that your data is excluded from any model training. Your data is processed exclusively within the Quentic environment and is never shared with or made accessible to other customers or external parties for model improvement purposes.

The Quentic team develops AI features using a combination of advanced language and logic models as well as specialized models for minor tasks such as audio transcription, translation, object detection, embeddings, and classification. For our model infrastructure, we rely on Azure AI, which provides granular control over deployment regions and data processing. This also includes specially trained and fine‑tuned models.

Yes. If required, you can deactivate all or selected AI features that are technically available in your Quentic system at any time. Quentic administrators can manage the relevant settings via a centralized AI control panel – ensuring your organization always retains full control over whether and which AI functions are used. Using AI is not a prerequisite for working with the Quentic Platform.

Quentic occupies both key roles under the EU AI Act: as a provider of AI features built into our platform, and as a deployer of AI tools in our own operations. We take both sets of obligations seriously. Our AI governance framework includes risk classification for every AI feature, structured development lifecycle requirements aligned with the Act’s provider obligations, and documented deployer controls for internal AI use. We have had prohibited-practice safeguards and AI literacy requirements in place since early 2025, in line with the Act’s first implementation phase.

Every AI feature in Quentic is designed with human oversight built in from the start. By default, AI in our platform suggests, summarizes, and accelerates – it does not make final decisions on your behalf. Where customers choose to configure automated actions, this is an explicit, controlled decision within their system – not something that happens without their knowledge. For safety-critical outputs such as hazard classifications, compliance assessments, or corrective action recommendations, Quentic’s AI provides supporting information that qualified professionals review and approve. This is not just a product principle – it is embedded in our governance policies and reflects our understanding that in EHS, the consequences of an unchecked AI error can be serious.

AI systems can generate outputs that sound confident but are factually wrong – a well-known phenomenon sometimes called hallucination. In an EHS context, this is not a minor inconvenience; it can be a safety risk. Quentic addresses this at multiple levels. Our AI features are designed to work within your dedicated Quentic environment, using your structured data as the primary input rather than relying solely on general-purpose models. Whenever AI generates suggestions or drafts, our platform clearly indicates which content is AI-generated, so users can apply appropriate scrutiny. Our governance requires that AI outputs involving safety, compliance, or regulatory content are verified against authoritative sources – not just checked for plausibility.

We use a risk-based approach aligned with the EU AI Act’s framework. Every AI feature in the Quentic Platform is classified early on during the design phase. Classification considers the type of data processed, the potential impact on end users, and any regulatory obligations that apply. Higher-risk features receive more rigorous oversight, documentation, and testing requirements. This tiered approach ensures that governance is proportionate: lightweight where the risk is low, rigorous where it matters most.

In the software interface, you can always clearly recognize which features are AI‑powered. We believe users should always know when they are interacting with AI, what data the AI is working with, and which steps in a process are automated. We are building Transparency Notes for our AI features – documentation that explains in plain language what each feature does, what data it uses, what its limitations are, and how human oversight is maintained. This reflects both our regulatory obligations under the EU AI Act and our belief that transparency is a prerequisite for trust – especially in regulated industries.

Yes. Quentic maintains a structured AI governance framework designed to be auditable. Our governance documentation is layered: a strategic framework with regulatory compliance mappings for auditors and leadership, and separate operational policies for day-to-day use. We maintain registries for approved internal AI tools and product AI features, each with documented risk classifications. Our existing ISO 27001 and ISO 9001 certifications provide the management system foundation, and our AI governance is designed for future alignment with ISO/IEC 42001, the international standard for AI management systems. We are happy to discuss our governance approach in detail during the procurement process.

For AI features within the Quentic Platform, our AI infrastructure is hosted on Azure AI, giving us control over deployment regions and data processing locations. AI processing of customer data occurs within the Quentic hosting environment and is not shared with other customers or external parties. We document data residency and processing locations for our product AI features, and for customers with specific data sovereignty requirements, we can provide detailed information about where and how AI processing takes place as part of the contracting process.