Cyberattacks on artificial intelligence (AI) systems are increasing, so it’s important users know their vulnerabilities and try to soften the damage if they get hit, according to a new report by the National Institute of Standards and Technology (NIST).



