Can You Trust Your Artificial Intelligence?

In January, Gustavo Sanchez, Pandata Tech, spoke to the chapter about mitigation and use cases of intrinsic risk for deployed artificial intelligence systems.

Cybersecurity is anything to do with digital risk. It is not exclusive to protecting systems from hackers and enemy bad actors. In terms of artificial intelligence (AI) and decision algorithms, there are two main cyber risks that need to be mitigated:

1. External Risk: This is what comes to mind whenever anyone hears 'cybersecurity.' Hackers and bad actors manipulating our systems.

2. Intrinsic Risk: Many think of a lot less of this when discussing cybersecurity. This includes flawed engineering, legacy equipment, bad data and what happens once external risk mitigation fails.

Although it may be counterintuitive, the number one risk to deployed AI systems is NOT bad actors and hackers, but the data the system depends on.

Event Photographs: