Seeing AI work changes the story
Morning Consult tested whether 65 real-world, positive examples of AI already at work — in medicine, fraud prevention, disaster response, and more — increase trust in AI. Respondents started skeptical. After reviewing the examples, trust moved substantially.
Trust in AI, before and after seeing the examples
"Overall, how much do you trust AI to do what is right?"
Before
After
Trust in AI companies, before and after
"Overall, how much do you trust AI companies (e.g., OpenAI, Google, Anthropic) to do what is right?"
Before
After
The five examples that moved trust most
Across 65 real-world AI examples tested, these produced the largest share of adults saying their trust in AI increased "a lot" or "somewhat."
Where AI earns trust
"Below are real examples of AI in use today. For each, how much does it increase your trust in AI, if at all?" — every example tested, organized into eight areas of real-world AI use. Each bar shows the share whose trust increased "a lot" (dark) or "somewhat" (light); the score is the combined total.
Where trust moves most, by audience
The single real-world AI example that increased trust the most within each group.
The awareness gap: biggest trust lift, least familiar
Examples where trust impact is high but most adults hadn't heard of them before this survey — the clearest opportunity for AI companies to tell an unfamiliar story.
What would most increase trust in AI companies
Single response, among all adults
How often adults use AI chatbots
ChatGPT, Gemini, Claude, Grok, Meta AI, and similar tools