Attorney and legal technology expert Colin S. Levy explains the scope of President Biden’s recent executive order on the development of artificial intelligence, and speculates about its long-term implications.

Biden’s executive order on ensuring the creation of “safe, secure and trustworthy” AI represents a “comprehensive, though not specific” attempt to address risks that might arise with the growth of the technology, Levy says. The order details a number of principles intended to address concerns such as bias, prejudice, public safety, national security and oversight of development.

The growth of AI in power, scope and influence has caused the U.S. and other countries to wrestle with the question of how to support innovation and technological advancements, while simultaneously ensuring safety. Some are calling for mandatory risk assessments before an AI system can be deployed, although it’s unclear at this point how that procedure would take place.

Included in the Biden order is the call for a “Bill of Rights” to safeguard civil liberties from the dangers of unchecked AI, but again, the details aren’t specified. Levy calls the order “a step in the right direction” toward creation of due process and encouragement of human review of algorithmic decisions. “AI is a set of tools that is best used in conjunction with human judgment and knowledge,” he says.

An unanswered question at this point is how the order would be enforced. Levy envisions some kind of permitting process, with specific punishments, including financial penalties, dictated for violations.

Also up in the air is how various government agencies — or, for that matter, a brand new agency dedicated to AI regulation — would be tasked with enforcing oversight, and whether legislative action will also be necessary to enact the necessary safeguards. “That will be challenging to do in a timely manner,” Levy says, adding that such efforts aren’t likely to keep pace with the rapid progress of the technology.