Andalibi: Can AI companies self regulate?

Amid fears of AI’s potential dangers, President Donald Trump and technology industry leaders signed a voluntary agreement on Sept.29 agreeing to oversee the safety of their own AI systems. 

As calls for government regulation continue, Trump described the agreement as “morally binding,” rather than legally enforceable. 

University of Michigan School of Information associate professor Nazanin Andalibi, an expert on the ethical implications of AI, told ABC News the voluntary agreement gives powerful companies too much control over defining AI’s harms and deciding how to address them.

“It lets the most powerful actors and people in the room define both the harm and also the remedy to the harm, and that's a strange model for achieving AI safety — unless safety isn't really the goal,” Andalibi says. 

RELATED 

Read “Trump says US AI giants agreed to 'police themselves' after White House meeting” from ABC news. 

Learn more about Nazanin Andalibi by visiting her UMSI faculty profile.