Kratsios was critical of the EU’s black-or-white segregation of ‘high-risk’ or not ‘high-risk’ AI.

At an interview hosted by the Hudson Institute, Michael Kratsios, the Chief Technology Officer of the US, commented on the White Paper released Wednesday by the European Union (EU) on artificial intelligence. (NextGov)

Heavy-handed compartmentalization

According to Kratsios, the EU had rather clumsily docketed AI technologies in the cubbyholes of “high-risk” or “not high-risk.”

He said that it appeared the EU would likely appoint a watchdog to determine the above classification of an AI technology and regulate it accordingly.

“I think this approach of bluntly bifurcating the entire AI ecosystem into two buckets is a little bit harsh,” Kratsios said, adding that the US approach would likely be more graded in its risk assessment compared to the EU.

“We think that AI regulations best serve on a spectrum, of sorts. There are certain kinds of AI-powered technologies that will require heavy regulatory scrutiny, and we in the United States are prepared to do that. But there are quite a few that need just a little, or not at all… I think creating this spectrum is important.”

He said that regulations should look to create a model that evaluates risk, use and is sector-specific. The model should have a graded spectrum and inbuilt flexibility to regulate appropriately for the risk attendant to each technology.

However, analysts who spoke to NextGov in the context of the Kratsios interview were more appreciative of the EU’s approach.

Still a work in progress

Aaron Cooper, vice president of global policy at BSA|The Software Alliance, felt that people were reading too much into the EU’s White Paper. It is not after all a regulatory framework.

Further, “it’s too early in both the development of AI policy in the EU and the US to know for sure whether they are really going to diverge,” Cooper said to NextGov.

Cooper added that it was not inappropriate for the EU to divide AI-related risk into buckets. “They are starting to move down the track towards trying to figure out what is the appropriate style of regulation.”

The US is light on protecting citizens and on data regulation

Agustin Huerta, vice president of technology, AI and process automation studios at Globant, was more critical of the US.

He appreciated that the EU had made progress after releasing guidelines last year. Further, its regulatory developments were more mature than those in the US.

“From the U.S. side, there seems to be too much focus on leading the industry compared to actually protecting citizens from biases or unethical uses of AI,” he observed.

Further, Agustin said the EU had more advanced data regulations compared to the US – an essential for achieving AI regulation.

“There is a need for the U.S. to catch up when it comes to data regulation. Some major flaws have occurred within the U.S. where several biases have been detected in AI algorithms.”

Related Story:    Artificial Intelligence: Europe Envisages Strict Rules for High-Risk AI