Blog

Welcome to the GIZ Data Lab Blog. Here you’ll find all the blog posts we’ve published, covering our use cases, activities, experiments, public communications, technical topics, and more. You can use the tag system to browse posts by topic. Many individual posts that don’t fit neatly into a more specific category are grouped under “Data for Development”. Enjoy exploring!

  • An abstract illustration of a deep neural network, with multiple connected layers of nodes in blue, green, and orange against a dark blue background. Four light-blue dots with a faint tail lead from the far left into the left-most column, representing input.
    Data Feminism

    Tan Minh Nguyen, Anna-Lisa Wirth |

    Are AI chatbots actually thinking?

    Why we should critically understand what AI models are really doing and why it changes how we should use them. 

  • An abstract geometric composition featuring black triangles and polygons on a pale blue background.
    Data for Development

    Data. Impact. Future.

    Data and artificial intelligence already are central to development cooperation and will play a major role in the future of GIZ. 

  • red greeting card reading "season's greetings 2025"
    General

    Anna-Lisa Wirth, Catherine Vogel, Robin Nowok, Teresa Kroesen, Vanessa Hochwald |

    Season’s Greetings from the GIZ Data Lab 2025

    Dear colleagues and friends, as the year comes to a close, as usual we wish to take a moment to reflect on our journey and share our plans for the upcoming year. In this time grappling with uncertainty, we have tackled new projects, connected with many remarkable people, and gained invaluable insights.


     

  • A woman with curly brown hair sits behind a desk, smiling at the camera, with a black laptop in front of her. In the foreground, a black bag hangs off the table, featuring a rainbow prism logo and the text "GIZ DATA LAB" in multicolored letters. Multiple papers, two water bottles, two microphones, and a mug are arranged on the desk.
    Data for Development

    Vanessa Hochwald |

    Farewell Vanessa

    After more than three and a half years, I am closing my chapter at the GIZ Data Lab. Looking back, time has flown by so fast, and yet it was filled with so many immensely gratifying, exciting, and humbling moments that it makes my head buzz.

  • Graphic featuring the text "NegotiateCOP" on a light blue gradient background, with fine white circular lines and a vibrant teal sphere on the right, and layered blue and teal waves along the bottom.
    AI in Negotiations

    Robin Nowok |

    NegotiateCOP

    Right on time for the 30th annual Climate Conference (COP30) in Belém, we launched our new tool “NegotiateCOP” that assists delegates navigate the extensive volume of documents submitted to the UNFCCC submission portal by more than 190 attending countries. 

  • Abstract geometric bird-like shapes in green, blue, pink, and light blue overlap on the left, while the right side shows an auditorium stage where a large group of people stand facing an audience. A large screen behind them reads "BECOND YEAR" in yellow letters, and two smaller screens display "Merck". In the foreground, an audience sits in an auditorium. Teal waves are across the bottom right.
    AI in Negotiations

    Anna-Lisa Wirth |

    NegotiateAI Goes PyCon

    Last April, PyCon DE 25 was bursting with exciting tech trends, and we had the pleasure of showcasing NegotiateAI, our negotiation AI designed to support the challenging plastic treaty negotiations at the United Nations. We walked the audience through our design process and design decisions for the Retrieval-Augmented Generation (RAG) system powering NegotiateAI, and the response was fantastic.

  • A red sketch of a bearded man with multiple corporate logos and fantastical figures. The sketch features logos for Google, Adobe, Nvidia, ChatGPT, and other tech companies.
    AI in Negotiations

    Yuri Irrgang, Anna-Lisa Wirth |

    Making Sense of AI Benchmarks

    Artificial intelligence is increasingly being integrated into professional workflows and organisational processes. However, as the industry boasts a myriad of model options, it has become difficult to know what individual models are capable of. Is there one specific AI that best suits my personal or professional needs?