The processing directory (VVT) acts as a database and supports decision-making processes. This article shows how companies can move away from a purely documentation-based perspective and integrate the VVT into a
Google will reclassify its role in the use of reCAPTCHA under data protection law and will no longer act as a controller but as a processor from April 2026. What this means
We present a practical 5-step plan that companies can use to bring AI governance to life. Each step highlights typical stumbling blocks and provides concrete recommendations for action that can be implemented through
More and more AI use cases are processing personal information, from customer data and employee details to sensitive analyses. The solution: AI governance and data protection compliance hand in hand. An integrated
For a long time, governance in the context of AI was seen as a necessary evil. Today, however, the picture is different: responsible AI is increasingly becoming a strategic success factor. Companies that
AI models are interchangeable, but their effects are not. Whether a system poses a low or high risk does not depend on the technology, but on the intended use.
Many risks do not arise in the cloud, but in the inventory. ITAsMa shows where systems are running, what data they process and when they were last checked.
From chatbots to scoring: every AI in a company needs measurable responsibility, clear evidence and a robust process. Those who take a structured approach now will protect data, brand value and
At the latest when companies implement several parallel AI projects, a common basis for evaluation is essential. Otherwise, different specialist areas, models and application types can hardly be compared with each other. A central
Model maps make AI comprehensible. They document the purpose, data origin, versions, performance metrics and limits of AI models. Those who use them consistently and integrate them into workflows gain auditability and