A study conducted by the Massachusetts Institute of Technology reveals a major flaw in artificial intelligence (AI): its inability to properly understand negation. This shortcoming could have dramatic consequences in critical sectors such as healthcare.