ERGO Group has made its internal generative AI platform, ERGO GPT, available to all employees after an early adopter phase and company-wide testing. The German insurer developed the large language model for internal use, with data protection and regulatory requirements built into the project.
ERGO said the decision to create its own AI environment was linked to the sensitivity of insurance data and the compliance standards applied across the group.
Employees were involved before the wider rollout. During the early adopter phase, staff tested ERGO GPT in daily work and identified additional use cases with internal teams.
Training was mandatory for participants. ERGO also collected employee feedback on how the system performed and where its functions needed adjustment before broader deployment.
ERGO GPT became available to all employees at the end of 2025. The insurer is using the system for tasks including structuring information, drafting and editing texts, preparing for meetings, developing ideas and handling repetitive administrative work.
The aim is to reduce time spent on routine activity and give employees more capacity for work requiring professional judgement or direct collaboration.
Christoph Thiel, CIO at ERGO, said employee response to the technology was as important as the software itself. He said staff had generally viewed AI as useful support rather than a replacement for existing roles.
ERGO’s approach combines an internal AI platform with employee training and ongoing feedback. For insurers, this type of deployment also addresses an issue surrounding enterprise AI adoption: staff need access to generative AI tools without moving sensitive information into uncontrolled public systems.
The group plans to continue developing ERGO GPT with its employees as adoption expands across different functions.
