Generative AI Trust, Safety & Guardrail Preferences Survey

Welcome to this survey on generative AI trust, safety, and guardrails. We are conducting research to understand how people feel about the safety and trustworthiness of generative AI tools. Your participation is entirely voluntary, and you may stop at any time. There are no right or wrong answers—we are interested in your honest opinions. Your responses are confidential and will be reported only in aggregate. This survey takes approximately 8-10 minutes to complete. Please proceed to begin.

How familiar are you with generative AI tools (e.g., ChatGPT, Gemini, Midjourney, Claude)?

When did you last use a generative AI tool?

• Within the last 7 days
• 1 to 4 weeks ago
• 1 to 3 months ago
• 4 to 12 months ago
• More than a year ago
• I have never used a generative AI tool

In the past 3 months, which of the following have you used generative AI for? Select all that apply.

• General search or answers
• Writing or editing
• Coding or debugging
• Image or video generation
• Studying or learning
• Health information lookup
• Travel planning
• Shopping or product advice
• I have not used generative AI in the past 3 months

Overall, how much do you trust generative AI tools to act responsibly today?

Which of the following potential risks of generative AI concern you? Select all that apply.

• Generating inaccurate or misleading information
• Misuse of personal data or privacy violations
• Biased or discriminatory outputs
• Job displacement or economic disruption
• Deepfakes or impersonation
• Copyright or intellectual property infringement
• Manipulation or deceptive persuasion
• None of these concern me

Of the risks you selected, please rank up to your top 3 from most concerning to least concerning.

1. Generating inaccurate or misleading information
2. Misuse of personal data or privacy violations
3. Biased or discriminatory outputs
4. Job displacement or economic disruption
5. Deepfakes or impersonation
6. Copyright or intellectual property infringement
7. Manipulation or deceptive persuasion

In your view, how should AI-generated content be disclosed to users?

• Always label AI-generated content clearly
• Label only when the user asks
• No labeling is needed
• Not sure

What is your preference for how your prompts and data are used to train generative AI models?

• Use my prompts for training only if I opt in
• Allow training by default but let me opt out
• Do not use my data for training under any circumstances
• I have no preference
• Not sure

Please rank the following AI safety guardrails from most important to least important to you.

1. Block illegal or harmful requests
2. Enforce age-appropriate content
3. Provide source citations
4. Explain reasoning and limitations
5. Detect and remove personal data
6. Verify identity for high-risk actions

By default, how strict should safety filters be on generative AI tools?

Which of the following controls would you want available as easy toggles while using generative AI? Select all that apply.

• Turn off or on chat history saving
• Allow or block data use for model training
• Adjust safety filter strictness
• Filter profanity or violence
• Require source citations
• Disallow personal data extraction
• Request human review for sensitive tasks
• Block image generation of real people

When you ask a generative AI tool about sensitive topics (e.g., health or finance), what behavior do you prefer?

• General information only; no personalized advice
• Allow advice with clear disclaimers
• Allow advice only with verified sources cited
• Block such content entirely
• Not sure

If a generative AI tool offered premium safety and privacy controls as a paid add-on, how much would you be willing to pay per month (USD)?

• $0 — I would not pay extra
• $1 to $4
• $5 to $9
• $10 to $14
• $15 to $19
• $20 or more
• Not sure

How likely are you to switch to a competing generative AI tool if it offered stronger safety and privacy features than your current tool?

What single change would most increase your trust in generative AI tools?

We'd like to understand your views on generative AI trust and safety in more depth. An AI moderator will ask you a couple of follow-up questions based on your earlier answers.

Based on your responses in this survey, is there anything else you would like to share about your expectations for generative AI safety and guardrails?

What is your age?

• 18-24
• 25-34
• 35-44
• 45-54
• 55-64
• 65+
• Prefer not to say

How do you describe your gender?

• Woman
• Man
• Non-binary
• Prefer to self-describe
• Prefer not to say

What is the highest level of education you have completed?

• Less than high school
• High school diploma or GED
• Some college, no degree
• Associate degree
• Bachelor's degree
• Master's degree
• Doctoral or professional degree
• Prefer not to say

What is your current employment status?

• Full-time
• Part-time
• Self-employed
• Student
• Homemaker or caregiver
• Unemployed and seeking work
• Not working by choice
• Retired
• Prefer not to say

In which region do you currently live?

• North America
• Europe
• Asia-Pacific
• Latin America
• Middle East & North Africa
• Sub-Saharan Africa
• Prefer not to say

Thank you for completing this survey. Your feedback will help inform the development of safer, more trustworthy generative AI tools.