Green Consumption Assistant: More sustainable shopping thanks to AI
The project “Green Consumption Assistant” supports consumers to quickly and easily discover resource-saving products and alternatives to new purchases online. The AI-supported assistant is researched and developed by a project network consisting of the Technical University of Berlin, the Berlin University of Applied Sciences and the green search engine Ecosia. It is funded by the German Federal Ministry for the Environment as a lighthouse project for artificial intelligence used for ecological challenges.
Inhumane working conditions in production and logistics, rising CO2 emissions and a high consumption of raw materials: although most German consumers are environmentally aware and feel responsible, these ecological and social impacts of consumption often fade into the background whilst shopping. The phenomenon – also known as the “knowledge-behaviour gap” – has a variety of different reasons. Crucially, consumers make most of their everyday consumption decisions on an ad hoc basis, but lack concrete and easily accessible information to identify sustainable products.

Image: CC license
This is why the Green Consumption Assistant project was launched in 2020. The AI-supported assistant can suggest more sustainable products and alternatives to new purchases to consumers during online shopping, so that these can be perceived just as spontaneously and easily as less sustainable purchase options.
Within the project, the interaction of three central building blocks is significant:
1. The development of an AI-based product database as central point of competence for sustainable consumption.
Despite rapid advances in the field of artificial intelligence (AI) for consumption recommendations, recommendation algorithms have not yet been successfully used to promote sustainable behaviour. We want to change that! With the help of AI and machine learning, a sustainability database will be created in which previously scattered and unstructured sustainability information on products will be aggregated, automatically categorised and finally, presented in a uniform format. The database will be continuously expanded and made available in open source to inspire further green applications.
2. Sustainable product options are made comparable and obtainable in real-time.
The collected information on products and services is made visible and comparable at the right moment via the recommendation assistant integrated into the search engine Ecosia. The system recognises whether the users are merely browsing the internet for information or have an intention to buy. If a purchase intention is recognised, general sustainability information of different product categories, but also sufficiency-oriented consumption options, such as sharing or buying second-hand, are brought to the consumers’ attention. With simplified presentations, our scientists ensure that sustainability factors are assessable for everyone.
3. An application design that does justice to users and their behaviour.
The intuitive usability of the application and the integration into existing infrastructures play just as central a role in the development of the GCA as the expectations of the users with regard to the quality and quantity of the content. Repeated tests and interviews with users combined with behavioural science serve as a basis for continuously improving the assistant. With this user-focused design, more sustainable action is made possible in all simplicity.
The project is funded by the German Federal Ministry for the Environment, Nature Conservation, Nuclear Safety and Consumer Protection as an “AI Lighthouse Project for the Environment, Climate, Nature and Resources” and aims to contribute to solving ecologically relevant challenges. Through its implementation on the search engine Ecosia, the “Green Consumption Assistant” offers the opportunity to help the multitude of scattered sustainability offers to achieve a significantly greater reach. The open-source development should also be understood as an offer for other search engines that would like to further develop in the direction of green search.