Can We Have “Trustworthy” AI?

Last week, a new framework for assessing AI information in health care was published (see The METRIC-Framework For Assessing Data Quality For Trustworthy AI In Medicine: A Systemic Review). The framework has five core elements, METRIC for short. These include measurement, timeliness, representativeness, informativeness, and consistency (METRIC). The standard considers such data qualities as understandability, consistency, credibility, and depth of the data. The authors noted that the standard is to assess the appropriateness of a dataset with respect to a specific use case rather than a one-size-fits-all approach. This is one of many efforts by health . . .