Popular AI Models Make Potentially Lethal Dosing Errors: Doximity Benchmark
· techtimes.com · popularity 18/100 · 0 similar stories today · AI & Data Science (Artificial Intelligence)
Read the original at techtimes.com →
Summary
AI medication dosing errors appear in 1 in 50 answers across nine widely used large language models, with 3.4% of mistakes classified as potentially lethal, according to a comprehensive Doximity benchmark released Thursday. The models failed most often at adjusting doses for kidney function, age, and drug interactions even when that patient data was explicitly provided.
Found in: Tech Times. Rating basis: single outlet; outlet tier 2 (unlisted); 1 feeds; feed position 10.