Learning Objectives
- Share AI safety knowledge effectively with different audiences
- Start productive conversations about AI risks without causing panic
- Help others build basic AI safety habits
Why Teaching Others Matters
You now know more about AI safety and privacy than most people. That knowledge is only fully valuable if you share it. The people around you — your family, your coworkers, your friends — are using AI tools with the same default settings and the same lack of awareness you had before starting this playbook.
You do not need to become an AI safety evangelist. But a few well-timed conversations can make a real difference in the lives of people you care about.
✅Tip
Lead with help, not fear. The most effective way to teach AI safety is to help someone do something specific — like changing a privacy setting — not to lecture them about risks. People remember actions they took, not warnings they heard.
Talking to Family Members
Your Partner or Spouse
Start with: "I just learned something interesting about ChatGPT — did you know it saves all your conversations by default? Here, let me show you how to turn that off."
This works because it is specific, helpful, and non-judgmental. You are offering to help with a concrete action, not delivering a lecture.
Follow up with: Share the privacy spectrum concept — low-risk versus high-risk information to share with AI. Most adults immediately understand the value of this framework.
Your Parents or Older Relatives
Older adults are often targeted by AI-enhanced scams — voice cloning, deepfake video calls, AI-generated phishing emails. Focus on:
- Voice cloning scams: "If you ever get a call from someone who sounds like me asking for money or personal information, hang up and call me back on my regular number. AI can now clone voices from just a few seconds of audio."
- AI phishing: "Scam emails are getting much more convincing because AI writes them now. Always check the sender's actual email address, not just the display name."
- Verification habit: "If something sounds too good or too urgent to be true, pause and verify before acting."
Your Children
See our AI for Parents & Families playbook for age-specific guidance. The key messages for kids:
- Never share your real name, school, address, or photos with AI tools
- AI can be wrong — always check important facts
- If AI says something that makes you uncomfortable, tell a parent
- It is okay to use AI for learning, but your homework should be your own work
Talking to Colleagues
The Casual Approach
At lunch or in a team chat: "I just changed my ChatGPT settings so my conversations are not used for training. Took 30 seconds. Want me to show you?"
This is low-pressure and helpful. Most people are curious enough to say yes.
The Formal Approach
If you want to advocate for team-level AI safety:
- Propose a 15-minute team discussion. Frame it as: "I want to make sure we are all on the same page about AI data handling, especially since we discuss client work in these tools."
- Share three specific recommendations: Turn off training data, do not paste client data into free tools, use anonymized scenarios when getting AI help with work problems.
- Offer to create a one-page guide. A simple document with platform-specific settings and do/don't guidelines. This is far more effective than a policy nobody reads.
💡Key Concept
The security champion model. Many companies have informal "security champions" — people who are not in IT but help their team stay aware of security practices. You can fill this role for AI safety without any formal authority. Just be the person who knows the answers and helps others.
If Your Company Lacks an AI Policy
This is an opportunity, not a problem. Consider:
- Raising it with your manager: "I think we should have basic guidelines for AI use. I have been researching this — would it be helpful if I drafted a one-page starter?"
- Connecting with IT or legal: They may already be working on something and would appreciate input from someone who actually uses the tools
- Starting small: A team-level agreement is easier to create than a company-wide policy
Teaching Moments to Watch For
The best teaching happens in natural moments, not scheduled lectures:
- When someone shares AI-generated content: "That is a great image — do you know if it is real or AI-generated? Here is how I check..."
- When someone pastes sensitive info into a chatbot: "Just a heads up — that tool might use your conversations for training. Want me to show you how to turn that off?"
- When a news story involves deepfakes or AI scams: "This is a good example of why I check sources before sharing. Here is what I look for..."
- When someone dismisses AI risks: "I used to think that too, until I learned that..."
⚠️Warning
Do not be the AI police. Nobody likes being corrected constantly. Pick your moments, offer help rather than criticism, and respect that some people will not want to engage. You can only share the knowledge — you cannot force people to act on it.
Creating a Ripple Effect
If every person who completes this playbook teaches three others the basics — privacy settings, verification habits, and the privacy spectrum — the impact multiplies rapidly. You do not need to teach everything. Just share:
- One setting to change (turn off training data)
- One habit to build (pause before typing sensitive info)
- One question to ask (can I verify this from another source?)
Three things. That is enough to make someone meaningfully safer.
Key Takeaways
- Your AI safety knowledge is most valuable when shared with others
- Lead with specific, helpful actions — not lectures or fear
- Different audiences need different approaches: family, colleagues, children
- Watch for natural teaching moments rather than scheduling formal lessons
- The ripple effect is powerful: teach three people three things and the impact multiplies
- Be helpful, not preachy — offer assistance, not criticism
