Human–AI Collaboration & Trustworthy Agentic Systems

KATLAS Technology is a UK SME developing operational governance infrastructure for human–AI collaboration and trustworthy agentic systems.

Our focus is the critical point between an AI recommendation and an authorised real-world action. In multi-party environments, AI may generate insight or propose an action, but organisations still need to know: who is authorised to act, what evidence is required, what constraints apply, when a human must intervene, and what proves the outcome afterwards.

KATLAS provides this operational layer through role-based authority, verifiable proofs, permissions, runtime decision gates, exception handling and signed operational receipts. This allows AI agents and human decision-makers to collaborate across organisational boundaries without requiring all underlying data to be centralised.

KATLAS can contribute:

• runtime governance for AI and agentic workflows;

• human-in-the-loop and human-on-the-loop decision models;

• role, authority and responsibility mapping;

• verifiable credentials and privacy-preserving proof exchange;

• policy and evidence gates before automated action;

• escalation, repair and exception workflows;

• signed and verifiable receipts for completed decisions and handoffs;

• distributed/federated operation across existing systems;

• sandbox demonstrators and pilot integration;

• operational evidence for assurance, evaluation and regulatory compliance.

Our demonstrators apply these principles to healthcare shared decisions, assisted mobility, freight and logistics, financial operations and industrial workflows.

For Horizon Europe, KATLAS is particularly interested in consortia working on human–AI collaboration, trustworthy and agentic AI, AI-enabled decision support, distributed data environments, robotics, autonomous systems and AI operating across multiple organisations.

We seek collaboration with AI and robotics researchers, universities, industrial operators, healthcare and mobility partners, system integrators, public authorities and consortium coordinators who need a practical governance and assurance layer between intelligent systems and real-world action.

KATLAS contributes the operational trust layer that helps turn AI capability into bounded, authorised and verifiable action.