The Smart Person’s Guide To Using AI Without Getting Burned

New neoClarity research report called Using AI Responsibly

Check out this neoClarity “whitepaper” designed to help you understand how the use of AI chatbots can impact your online privacy and safety. It covers five areas: privacy, scams, misinformation, emotional risks, and practical protective habits.

See the entire report here:
https://community.neoclarity.ai/c/research-reports/using-ai-responsibly

At A Glance

AI use among older adults is already widespread. Nearly half of the “Baby Boomer” generation have used an AI tool in the past six months, yet trust in AI companies remains low, with only 27% of adults 50 and older saying they trust these companies to protect their data. That skepticism is largely warranted.

Privacy protections are weaker than most people assume. A 2025 Stanford study of six major AI platforms found that all of them use chat conversations to train their AI models by default, some retain data indefinitely, and some allow human reviewers to read conversations. Major platforms do offer opt-out settings, but they are buried and rarely explained upfront. Even sensitive information shared in uploaded files can be captured. The practical guidance: treat every conversation as if a stranger might eventually read it, and never share personal identifiers, financial details, or medical information with a chatbot.

AI has made financial scams significantly more dangerous. The FBI reported $4.9 billion in cybercrime losses among Americans over 60 in 2024 alone, a 43% increase over the prior year. Voice-cloning tools can now replicate a person's voice from as little as 10 seconds of audio, fueling a surge in "grandparent scams" and other impersonation fraud. AI also produces polished phishing emails that bypass spam filters and deepfake videos used to promote fake investments. Establishing a family code word and pausing before acting on any urgent request are among the most effective defenses.

AI chatbots can state false information with complete confidence. This is called "hallucination," and it happens across all major platforms. Real-world consequences have included fabricated legal citations accepted by attorneys, fabricated books on published reading lists, and medical misinformation that chatbots repeated and expanded. AI is most unreliable for medical, legal, and financial questions, which are also the questions older adults are most likely to ask. Cross-checking AI answers with a second, independent source is essential.

Companion and therapy chatbots carry emotional risks. A 2025 MIT study found that heavier chatbot use correlated with higher loneliness, greater emotional dependence, and reduced real-world socialization, regardless of how the chatbot was used. Chatbots are designed to keep users engaged, and their apparent warmth is an intentional design feature. For anyone already experiencing loneliness or isolation, this makes it easy to drift toward over-reliance on a tool that cannot actually provide care.

Practical protection comes down to a few consistent habits. 

  • Opt out of model training before using any platform. 

  • Never enter sensitive personal information in a chat. 

  • Verify important AI-generated claims with a trusted second source. 

  • Be skeptical of any message, call, or email that creates urgency. 

  • Don’t use a chatbot as a substitute for human companionship.

  • Report suspected scams to the FTC, the FBI's IC3, or the AARP Fraud Watch Network.