Health, money, and law are exactly where an ordinary person feels most outmatched β and so exactly where AI is most tempting to lean on. It genuinely can help with all three. But these are also where a fluent wrong answer does the most damage, and where AI is most prone to inventing specifics. The rule that keeps you safe is the same in all three: use it to understand and prepare β never to diagnose, decide, or stand in for the professional.
Health
β Great for
Explaining a diagnosis, test result, or medication in plain words; preparing questions for your doctor; understanding what a treatment generally involves; getting a sense of whether something is worth a visit.
β Dangerous for
Diagnosing yourself, dosing, or "is this an emergency?" It can miss red flags and state wrong doses with total confidence. It hasn't examined you and doesn't know your history.
β Safe move
Use it to prepare for and understand the appointment, not instead of one. Never start, stop, or change medication on its word. Anything severe, sudden, or worsening β contact a clinician or emergency services now.
Money
β Great for
Explaining a financial product, a fee, or a statement; defining terms; understanding options and trade-offs; drafting a dispute or a budget you then check; the questions to ask an advisor.
β Dangerous for
Specific investment, tax, or benefits advice; the actual numbers in your case (it can be confidently wrong); anything that depends on current rates or rules; "should I buy or sell?" Confident, generic answers dressed up as if they fit your case.
β Safe move
Use it to understand and prepare; verify every number yourself; and for big, tax-related, or irreversible money decisions, confirm with a qualified advisor or the official source before acting.
Legal
β Great for
Explaining a clause, a notice, or legal jargon; understanding the general process and your options; drafting a letter or a list of questions; figuring out what kind of help you actually need.
β Dangerous for
Citations and case law β it invents them, and people have been sanctioned for filing fake AI-generated cases. Also jurisdiction-specific rules, deadlines, and acting as your lawyer. Law varies enormously by place.
β Safe move
Use it to understand and to draft β never to cite or to replace advice. Verify any law or citation against an official source, mind your local rules and deadlines, and for anything consequential get a real lawyer or legal aid (often free).
True for all three
The close
Used right, AI is genuinely leveling here β it can translate the intimidating language of medicine, money, and law that used to require paying a professional just to understand. That's real power in an ordinary person's hands. The discipline that keeps it a gift rather than a trap is one sentence: understand and prepare with the AI, decide and verify with a human. Before you rely on any answer in these areas, run it through the trust check β and when it's your health, your money, or your rights, treat "the AI said so" as the start of the work, never the end of it.