Coffee Break #3: Using AI for Report Writing

HOST

Youth Agency

DATE

May 21, 2026

TOPIC

Using AI for Report Writing

Description

Writing reports is part of everyday work in many youth organisations, which it can also be time-consuming and repetitive.

In this AI Coffee Break, we will explore how AI tools that can support you in structuring information, summarising content and drafting reports more efficiently with real-life example and practical workflows.

This session is designed for youth workers, youth information workers, and anyone working with young people who would like a simple introduction to AI. No prior experience with AI is required.

Link to the meeting

Highlights

  • AI won’t write your report for you, and it shouldn’t. The workflow presented uses AI as a drafting assistant, with human proofreading and fact-checking as a non-negotiable final step.
  • Two tools, one chain: NotebookLM plus a chatbot. NotebookLM answers only from the sources you upload and shows where it found each fact, then ChatGPT or Gemini condenses that answer to the report’s character limit.
  • Let AI write your prompt. ChatGPT produces a far better prompt than most users would write themselves, including instructions to stay source-based, avoid speculation, and flag missing information instead of inventing it.
  • Using AI in Erasmus+ reports is allowed. The German National Agency confirmed there is no prohibition, and the European Commission has published guidance saying the same: use it, but personalize it and know what you wrote.
  • If it sounds like AI to you, it sounds like AI to the reviewer. Tone matters more than tooling, and an unedited generated text is immediately recognizable to assessors.
  • Protect your data. Never upload sensitive information such as bank details or birth dates, strip those pages from PDFs first, and use a paid or organization-licensed version for work data.
  • Gemini pairs better with NotebookLM than ChatGPT does. Same company means a direct connection, no copy-pasting, and no re-uploading documents, which also saves tokens.
  • NotebookLM takes more than PDFs. It processes Word documents, audio recordings, video, and YouTube links, and connecting Google Drive keeps sources updating live.
  • Build a prompt library. Keep a dedicated notebook containing your most useful prompts and reuse it as a database instead of rewriting them each time.
  • Human in the loop, always. AI sees a project only from the technical side and drops details that matter subjectively to the team, so shortened texts need checking for accuracy and omissions.