On October 15, according to the British Guardian, 13, several experts warnedAI Application in the medical field may raise complex questions of legal liability, and when medical errors occur, the determination of liability may become confused。
CLINICAL AI HAS DEVELOPED RAPIDLY, AND RESEARCHERS HAVE DEVELOPED A VARIETY OF TOOLS, RANGING FROM ALGORITHMS THAT AID IMAGE INTERPRETATION TO SYSTEMS THAT SUPPORT DIAGNOSIS, TO AI. AI IS ALSO BEING APPLIED TO HOSPITAL MANAGEMENT, FROM OPTIMIZING THE USE OF BEDS TO IMPROVING SUPPLY CHAIN EFFICIENCY。
EXPERTS POINTED OUT THAT AI HAD CONCERNS ABOUT THE MULTIPLE BENEFITS THAT WERE EXPECTED, INCLUDING:The validity of the tool has not yet been fully validated,as well asIt is difficult to define the attribution of responsibility in the event of adverse results.
PROFESSOR DEREK ANGUS OF THE UNIVERSITY OF PITTSBURGH SAID: “THERE MUST BE A SITUATION IN WHICH A PROBLEM IS PERCEIVED AND AN ATTEMPT IS MADE TO FIND THE RESPONSIBLE PARTY.” ANGUS WAS THE FIRST AUTHOR OF THE REPORT OF THE AI SUMMIT ORGANIZED LAST YEAR BY THE AMERICAN MEDICAL ASSOCIATION MAGAZINE. THE REPORT ANALYSES NOT ONLY THE CHARACTERISTICS AND AREAS OF APPLICATION OF THE AI TOOL, BUT ALSO THE LEGAL CHALLENGES IT POSES。
PROFESSOR GLENN COHEN OF HARVARD LAW SCHOOL, CO-AUTHOR OF THE REPORT, STATED THAT PATIENTS MAY EXPERIENCE DIFFICULTIES IN PROVING FAULT IN THE USE OR DESIGN OF AI PRODUCTS. ON THE ONE HANDHard to get in-house running information on the toolOn the other handIT IS DIFFICULT TO PROPOSE A REASONABLE ALTERNATIVE DESIGN OR TO PROVE THAT THE NEGATIVE RESULT IS CAUSED BY AII don't know. According to Professor Cohen: “The parties may abdicate each other's responsibility and may, by contract, redistribute responsibility or involve litigation for damages, which may make it more difficult to bring an action.”
PROFESSOR MICHEL MELO OF STANFORD LAW SCHOOL, WHO WAS ONE OF THE AUTHORS OF THE REPORT, SAID THAT THE COURTS WERE ABLE TO RESOLVE RELEVANT LEGAL ISSUES, BUT THAT EARLY PROCESSING WAS TIME-CONSUMING AND UNCERTAIN, INCREASING THE OVERALL COST OF AI INNOVATION AND APPLICATION OF ECOSYSTEMS. “THIS UNCERTAINTY INCREASES THE COST OF AI INNOVATION AND APPLICATION.”
ReportCONCERNS WERE RAISED ABOUT THE EVALUATION OF THE AI TOOLIT WAS NOTED THAT MANY TOOLS WERE NOT MONITORED BY REGULATORY BODIES SUCH AS THE UNITED STATES FOOD AND DRUG REGULATORY AUTHORITY. ANGUS STATED: “CLINICIANS ARE CONCERNED WITH IMPROVING HEALTH OUTCOMES, BUT REGULATORS DO NOT NECESSARILY REQUIRE PROOF. ONCE IN USE, AI TOOLS MAY APPEAR IN DIFFERENT CLINICAL SETTINGS, FACE DIFFERENT PATIENTS AND BE OPERATED BY USERS WITH DIFFERENT SKILL LEVELSUnforeseeable situationsI don't know. The practical use of seemingly feasible tools at the approval stage may vary considerably.”
1AI WAS INFORMED FROM THE REPORT THAT THE REPORT ALSO INDICATES THAT THERE ARE MULTIPLE BARRIERS TO THE ASSESSMENT OF THE AI TOOLS, SUCH ASIt usually takes clinical use to make a full assessmentTHE CURRENT ASSESSMENT IS EXPENSIVE AND CUMBERSOME. ANGUS EMPHASIZED THE NEED TO INVEST IN A FULL ASSESSMENT OF THE AI TOOLS IN HEALTH CARE AND THE FOCUS ON BUILDING DIGITAL INFRASTRUCTURE. “THE SUMMIT DISCUSSIONS FOUND THAT THE MOST ADEQUATE TOOLS ARE OFTEN THE LEAST USED AND THE MOST WIDELY AVAILABLE TOOLS ARE THE LEAST EVALUATED.”