← Home
PLEN

AI Transparency

Last updated: 11.06.2026

1. How we use artificial intelligence

Guardiso uses artificial intelligence as a tool that supports work on security and compliance — for example to draft policies, help answer security questionnaires, analyse risks and find information across your documentation. AI in Guardiso is there to assist people, not to replace their decisions.

This notice reflects the transparency principles set out in Regulation (EU) 2024/1689 of the European Parliament and of the Council (the AI Act). It describes how, for what purpose and with what safeguards we use AI.

2. A human always decides (human-in-the-loop)

Content prepared with AI assistance is a draft. The final decision — approving a policy, assessing a risk, answering an auditor or a customer — is always made by a person. The platform does not make automated decisions producing legal or similarly significant effects for you or your employees without human involvement (in line with Article 22 GDPR).

3. Labelling AI-generated content

Where content is created with AI, we mark it as AI-assisted and, where possible, point to the sources in your documentation the assistant relies on. This way you always know which content needs your review before approval.

4. AI models and data flow

We use third-party large language models as subprocessors, under data processing agreements (DPAs) and Standard Contractual Clauses (SCCs):

  • Anthropic — assistant and analysis. Only the query content is sent to the provider, data is processed transiently and is not used to train models.
  • OpenAI — generation of embedding vectors for semantic search. Customer data is not used to train models.

Your database, files and personal data are not sent to AI providers beyond what is necessary to perform a given request. The full data-flow description is available on our subprocessors page.

5. Optional features and their limitations

AI features are optional — the platform can be used without them. Please note that AI models can sometimes produce inaccurate or incomplete content. Treat every output as working material to be verified, not as finished advice. Guardiso is a supporting tool and does not replace legal advice, a certification audit or the decisions of qualified professionals.

6. Contact and questions

Questions about how we use AI? Contact us:

See also: