The UK’s Department for Transport (DfT) is using Google Cloud’s AI tools to dramatically speed up the way it analyzes public consultation responses. According to Google’s official blog, the department handles around 55 public consultations every year, many of which generate more than 100,000 free-text responses from citizens. Traditionally, reviewing and categorizing this feedback took several months of manual work, creating delays in publishing policy responses.
To solve this, the DfT partnered with Google Cloud and the Alan Turing Institute to build the Consultation Analysis Tool (CAT) on Vertex AI using Gemini models. The system can now identify themes, classify responses, and extract insights in just a few hours instead of months. Google states that the tool has achieved up to 90% accuracy in evaluations and could help save the department up to £4 million annually.
The AI rollout goes beyond consultation analysis. The DfT is also using Google Cloud services such as Cloud Run, Cloud CDN, and Firestore to support tools for urban planning and infrastructure decisions, including a connectivity platform for sustainable transport planning. Another system, the AI Correspondence Drafter, helps create first drafts of responses to public inquiries by pulling information from secure policy databases and using Gemini to draft replies.
Overall, the article highlights how AI is increasingly being used to modernize public administration and policy-making. A key point is that the DfT is following a human-in-the-loop approach, where policy experts review outputs for fairness, accuracy, and bias before decisions are made. This positions AI as a tool to improve government efficiency while keeping human judgment central to public policy.
