ENERATIVE AI
AI advice grows in popularity, despite the risk
ChatGPT has rapidly become one of the most popular sources for advice in America, but overreliance on AI could be leading people astray.
A new report from Pearl.AI found that 40% of 2,000 Americans surveyed use AI weekly, looking for advice on everything from medical issues to financial and legal questions.
Over the past 6 months, 65% of respondents said they’ve used generative AI for issues they previously trusted only to human experts.
However, the report found that a large number are being misled.
The findings showed:
- 22% of Americans have followed AI’s medical advice, which was later proven wrong
- 42% of Millennials believe AI can give them all the financial advice they’d ever need
- 19% have lost money from bad AI advice
- 28% of Americans would sign a legal document drafted entirely by AI
- 31% would let an AI lawyer defend them in court
Pearl founder and CEO Andy Kurtzig said the trend stems from cost and accessibility barriers faced by the general public, particularly those in urban communities.
He said, however, that turning to AI as an alternative resource is a “dangerous gamble.”
“The promise of AI is speed, but its defining weakness is confidence without certainty, Kurtzig told The Deep View.
“We’re being sold a tool that mimics authority it hasn’t earned, creating a structural safety gap in every high-stakes field it touches,” he said. “The risk isn’t just bad information; it’s the illusion of expertise.”
The response, Kurtzig said, should be to maintain a human in the loop strategy when it comes to building AI systems through a “hybrid intelligence” blending AI’s accessibility with “the indispensable wisdom of a verified human expert.”