BlueCallom AI Policy
1) ARTIFICIAL INTELLIGENCE USAGE
BlueCallom uses Artificial Intelligence as a foundational technology across all our products. We also use AI on our website, especially in our chatbot Mentoha.
2) REGULATORY FRAMEWORK SUPPORT
We are supporting the following regulatory frameworks:
-
EU AI Act: The world’s first comprehensive AI law, the EU AI Act uses a risk-based approach.
-
Unacceptable Risk: Systems like social scoring or real-time biometric identification in public spaces are banned as of February 2025.
-
High-Risk: Systems used in critical sectors (e.g., healthcare, hiring) must meet strict safety, transparency, and data quality standards.
-
General-Purpose AI (GPAI): Large models (like those behind chatbots) became subject to transparency and copyright rules in August 2025
-
We are using BlueCallom COMPLY-AI to track changes and new products, as well as the respective workflows
-
-
GDPR (General Data Protection Regulation): While not AI-specific, it governs how AI models process personal data, requiring explicit consent and providing individuals a “right to explanation” for automated decisions.
3) TRANSPARENCY & ACCOUNTABILITY
BlueCallom users know they are interacting with AI when using any of our Applications / Flights.
We are providing training to help people understand the functionality of the AI solution we provide.
Within our organization, the responsibility for the AI’s outcomes and establishing mechanisms for review and updates.
4) PROTECTED AI APPLICATION DESIGN
BlueCallom AI APPLICATIONS and FLIGHTS are built as complete and ready-to-run Enterprise AI Solutions. Unlike individual or self-made agents, where anybody can modify their functionality, our Application design is built for a protected workflow. Functionality and data access are carefully protected and cannot be accessed by a user. Business users would not be asked or allowed to use data that the system doesn’t expect.
5) HUMAN IN THE LOOP
We fully support the idea of a human-in-the-loop. All autonomous actions that have decision-making characteristics have a human-in-the-loop, yet a customer can turn off each step individually in the setup of each Application or Flight.
6) AI ASSET APPROVAL
The BlueCallom Platform provides an organization-level approval for any AI Asset (Prompt, Agent, Application, Flight) to be used by its respective users. Unauthorized AI Usage is nearly impossible and would be tracked regardless. This reduces the risk of shadow AI.
In BlueCallom·AI, there are different forms of autonomous workflows. These are not introduced silently or added later without transparency. Instead, they are explained, configured, and consciously confirmed with the customer as part of the setup and before operational use.
7) AUTONOMOUS WORKFLOW MANAGEMENT
In BlueCallom·AI, there are different forms of autonomous workflows. These are not introduced silently or added later without transparency. Instead, they are explained, configured, and consciously confirmed with the customer as part of the setup and before operational use.
8) BLUECALLOM RESPONSIBILITY
BlueCallom’s responsibility is to ensure that the AI does only what it is intended to do — and only what it is technically allowed to do — within the defined task scope, description, and approved configuration. Autonomy at BlueCallom, therefore, does not mean uncontrolled action. It means clearly described, limited, and traceable execution within a governance framework defined by the customer.
We therefore understand Corporate AI not merely as the use of technology, but as a governed extension of operational business processes. Ownership, control logic, intervention rights, and management responsibility are not treated as external considerations. They are an integral part of the BlueCallom approach to safe, regulated, and accountable AI deployment.
The customer can define the degree of autonomy for each agent at any time. A fully autonomous agent can be configured so that, at defined decision points, a “Human in the Loop” is involved. In that case, responsibility for review, approval, or intervention is assigned to a specific person or role.
As a general principle, our multi-agent systems and AI applications are set up, described, and activated by the customer. This means that steering, control, and domain responsibility remain with the customer. BlueCallom provides the technical and methodological environment in which these responsibilities can be transparently defined and operationalized.
Each user must declare that using BlueCallom means also using AI. Everything that can be done with BlueCallom is AI-based. However, a user cannot upload or inject AI applications, nor can they access the BlueCallom Development system without authorization by a respective manager.
11) AI MODEL TRAINING POLICY
We do not use any data obtained from ‘Data Transfer Providers’, including Google Workspace APIs, to develop, train, or improve generalized (non-personalized) AI or ML models.
12) PRIVACY
Updated: May 4, 2026