Malpractice Liability When Doctors Use AI Quiz

Answers and explanations appear as you take the quiz.

1. Which set lists the four elements a US malpractice claim usually requires?
   • Intent, motive, opportunity, harm
   • Duty, breach of the standard of care, causation, damages
   • Consent, disclosure, billing, harm
   • License, specialty, error, apology
2. Under the 2019 JAMA analysis by Price, Gerke and Cohen, which scenario carries the highest liability risk for a physician?
   • Following an AI recommendation for standard care
   • Rejecting a correct but nonstandard AI suggestion
   • Following an AI recommendation that departs from standard care when the patient is then harmed
   • Not using AI in a specialty where no tools exist
3. Why can an undocumented dismissal of an AI flag hurt a clinician in court?
   • EHR audit logs show the alert was displayed, which suggests the clinician was on notice
   • AI companies testify against doctors
   • Dismissing alerts is a crime
   • The flag automatically becomes the standard of care
4. Why is FDA 510(k) clearance not a full shield for a physician using an AI tool?
   • The FDA does not review any AI software
   • Clearance expires after one year
   • Clearance makes the FDA the defendant
   • 510(k) clearance generally does not block state negligence claims or define the standard of care
5. A hospital deploys a model without validation or training, and a patient is harmed. Which theory might a plaintiff use against the hospital itself?
   • Learned intermediary
   • Corporate negligence over how the tool was selected, validated and rolled out
   • Product recall
   • Informed refusal
6. Why are product liability claims against AI developers harder to win, according to the guide?
   • Developers are immune under federal law
   • Software never causes harm
   • Courts often treat software as a service, and the clinician stands between the tool and the patient as a learned intermediary
   • Patients cannot identify the developer
7. Which documentation practice does the guide recommend when AI informs a clinical decision?
   • Record the tool used, its output and why you agreed or disagreed based on your own assessment
   • Never mention AI in the chart
   • Copy the vendor's marketing claims into the note
   • Record only the final diagnosis
8. An ambient AI scribe inserts an exam finding that never happened, and the physician signs the note. Who is responsible for the note's accuracy?
   • The scribe vendor alone
   • The patient
   • No one, because the error was automated
   • The physician who signed it