Module Check: Why Safety Matters
You have traveled the full arc of this module. You started by defining AI safety and understanding why powerful tools demand proportional care. You learned the five major ways AI can go wrong. You distinguished accidental failures from deliberate misuse. You built a framework for assessing stakes. You mapped the ecosystem of people and institutions responsible for safety. You examined safe habits for everyday AI use. And you practiced reasoning about real applications in the Stakes Sort. This final lesson is not a repeat — it is a consolidation. The goal is to make sure all these ideas are connected in your mind, ready to apply in the next module and in your life.
Key Terms Review
Flashcards — click each card to reveal the answer
Module Quiz
A city deploys an AI system that uses traffic camera footage to automatically issue speeding tickets, with no human review of individual cases. Using the stakes framework, what is the strongest argument that this is a high-stakes application?
An AI language model writes a history essay for a student. The essay includes three specific dates and two quoted sentences attributed to a famous historical figure. The student submits it without checking. What is the primary risk?
A company's AI hiring screener consistently ranks candidates from a particular university lower than equally qualified candidates from other universities — a pattern that was present in the historical hiring data it trained on. Who bears primary responsibility for fixing this?
Which response is most appropriate when someone you know is in a mental health crisis and reaches out to you?
Regulators in the European Union decided that AI systems used for medical diagnosis must meet much stricter requirements than AI systems used for playlist recommendations. What principle does this reflect?
A social media platform's AI recommends increasingly extreme political content to users because extreme content generates the most comments and shares. Separately, a scammer uses an AI voice cloner to impersonate a user's relative in a phone call. Which of these is an accidental failure and which is deliberate misuse?
AI safety is not about fear — it is about care. Powerful tools demand proportional responsibility. Understanding the ways AI can go wrong, who is accountable, and how to act wisely as a user makes you a participant in getting AI right, not just a bystander.
Synthesis: Design a Safety Briefing
- Imagine you are an AI safety advisor asked to brief a group of adults — parents, school administrators, or city council members — who use AI tools regularly but have never studied AI safety.
- Step 1: Choose your audience (parents, administrators, or city council members).
- Step 2: Select the three most important ideas from this module to include in a five-minute briefing. For each idea, write two to three sentences in plain language — no jargon — that explain why it matters to your specific audience.
- Step 3: Choose one real AI application your audience likely uses (a school management system, a city service, a parenting app) and apply the stakes framework to it. Rate it on severity, reversibility, and scale, and recommend one specific safeguard.
- Step 4: End your briefing with one concrete action you would ask your audience to take in the next week to make their use of AI safer.
- Step 5: Present your briefing to a partner or small group. Ask for feedback: Was it clear? Was the most important idea actually the most important for that audience? Did the recommended action feel realistic?