Commission launches a new general-purpose AI tool - GPT@EC

The tool has been developed by the Directorate-General for Digital Services (DIGIT), and is based on a successful previous project, GPT@JRC, that the Joint Research Centre (JRC) has put in place for generative-AI scientific and research purposes. GPT@EC offers an alternative to third-party generative AI tools, limiting the risks of mishandling Commission data.  

“I am confident that GPT@EC will improve the efficiency and the effectiveness of our internal administration, enhance our productivity and responsiveness, and help us to better deliver our services. Above all, it will allow us to do our job and serve EU citizens better, all while knowing that we control access to Commission information.” 

Johannes Hahn, Commissioner for Budget and Administration 

The GPT@EC tool will enable staff to enhance their productivity and creativity by assisting them in tasks such as drafting texts, summarising documents, developing software code and more, while keeping their data secure. The tool adheres to corporate rules and compliance standards. Its big advantage over the third-party tools is that it is adapted to the Commission context, allowing the processing of the internal Commission information, without sharing it with external parties. GPT@EC focuses on supporting staff with a wider variety of tasks, by providing access to several Large Language Models (LLMs), each suitable for different tasks and sensitivity levels. It also offers the possibility to use the open source and sovereign models. 

“We want Commission staff to benefit from all the possibilities that AI has to offer, but we need to give them the tools and guidance to do it safely. This is where GPT@EC steps in. We have developed it as part of the efforts to ensure the safe use of AI in the Commission. The feedback of our staff is now key to further improving the tool and making it a better fit for the needs of the Commission.” 

Veronica Gaffey, Director-General for Digital Services 

The roll-out of GPT@EC is fully aligned with the goals of the AI@EC Communication adopted in January this year, the Commission’s internal vision for the development and use of lawful, safe and trustworthy AI. The launch of new AI-based tools aims to accelerate the digital transformation of the Commission’s departments and executive agencies.