Data And Bias
Data is "the issue of our age," Rometty said.
But the increasing significance of data across the business landscape raises all sorts of complicated questions about how tech companies manage and protect that unique asset for their customers, and how they help customers formulate decisions by applying artificial intelligence.
To allow AI to thrive, companies that develop the technology must make transparent to the public what data sets their algorithms have been trained on, and who trained them, Rometty said.
"You can train bias if you're not careful," she warned.