Responsible AI for Public Work

English version (about 1 min 50) 0:00 · On screen: A signed memo. Invented figures. Imagine this. A manager asks the AI to write a memo. The text is perfect. But the figures are wrong: the AI made them up. The memo is signed, and nobody notices, until a colleague checks. 0:20 · On screen: Module 1 — Understanding AI and using it responsibly Welcome to Module 1. You will see what AI can do, how to protect citizens' data, and what Beninese law says. 0:35 · On screen: 1. The AI suggests. You check. You decide. AI is very useful. It writes a first draft, rewords, summarizes and translates. But it can make up a number or a law, with the same confidence as when it tells the truth. This is called a "hallucination". The golden rule: the AI suggests, you check and you decide. 1:00 · On screen: 2. Anonymize before you paste Next, protect citizens. A text pasted into a free tool leaves your office. First, replace names, file numbers and addresses with neutral words, like "Mr. A". Never share health, court or national security data. 1:25 · On screen: 3. Digital Code, Book 5 · APDP Finally, the law. In Benin, the Digital Code, Book 5, protects personal data. The APDP oversees it, and any citizen can file a complaint with it. The European GDPR is stricter, but the spirit is close. And "the AI wrote it" excuses nothing. 1:45 · On screen: To remember — the responsibility stays with you. To remember: AI helps, but the responsibility stays with you. Now, go to the first lesson.