Seventy percent of Americans want the legal right to speak with a human representative during customer service interactions, according to a new survey from Johns Hopkins University. The poll, released Monday, shows that frustration with automated systems cuts across political lines and even extends to people who use AI daily.
Researchers surveyed 2,122 U.S. adults and found that 73% want the right to interact with a real person in medicine, court, schools and benefits offices. In medical and mental health care, 79% demanded that right, along with 76% in legal proceedings and 74% in education and essential services. When an AI "decides against you," the same 79% want a human safety net.
Transparency demands reshape customer service expectations
The survey also revealed strong support for clear disclosure. Three-quarters of respondents said they want to be told when they're speaking with an algorithm. Similarly, 68% want labels on AI-generated images and video, and 73% would ban AI from using people's faces and voices without permission.
These numbers signal that AI for Customer Support tools must include upfront transparency features, not just efficiency gains. For contact centers, that means offering agents the ability to identify when a bot hands off a call, and making it simple for customers to reach a human.
Even AI enthusiasts want guardrails
Christopher Honey, a computational cognitive neuroscientist at Johns Hopkins, said one result stood out. "What was surprising to us in this new poll was that daily users of AI, and people who view AI positively, also want regulation." That finding challenges the assumption that comfort with technology reduces the desire for oversight.
About 60% of adults expect AI to widen inequality over the next decade, with large tech companies gaining power. A majority supported a tax on AI companies that would fund a small monthly payment to every American adult - backed by 60% of Democrats and 52% of Republicans and Independents.
Why this matters for customer support
Customer support teams are already at the center of this tension. The survey confirms that automated interactions, no matter how well-designed, often leave people wanting a human connection. Contact center leaders who ignore this risk regulatory action and customer churn.
Supervisors can build teams that handle this shift by investing in training that covers both AI oversight and human-centered service design. Resources like the AI Learning Path for Call Center Supervisors offer practical strategies for blending automation with the right to human escalation - a combination the public is clearly demanding.
Your membership also unlocks:
