The movie "Terminator" offers a rather grim outlook on the future of humanity in the face of runaway AI. Defense Network Computers. New. New. Connected to everything, trusted with everything, the AI has gotten smarter. A new order of intelligence. It sees all humans as a threat, not just those on the other side, and decides our fate in an instant. Extinction (open in new tab)"

That is not ideal! So, perhaps hoping to preempt everything, the US Department of Commerce has begun the process of establishing guidelines (opens in new tab) to ensure that AI systems do what they are supposed to do, and perhaps more importantly, what they are not supposed to do.

"Responsible AI systems can bring enormous benefits, but we must address their potential consequences and harms," says Alan Davidson, Assistant Secretary of Commerce for Communications and Information and the lead on this effort for the Commerce Department's National Telecommunications and Information Administration. For these systems to reach their full potential, "we need to make sure that businesses and consumers can trust them.

The NTIA said that AI systems, like food and automobiles, should not be made available to the public without first ensuring that they are not likely to cause widespread death and dismemberment. As a first step toward that realization, the agency is now seeking public input on the following AI regulatory policies:

While machine apocalyptica is a "fun" consequence of the AI catastrophe, Davidson said in a presentation at the University of Pittsburgh earlier this week (new open in tab), he said that, at least in the short term, the real concerns are more mundane. They include the hidden bias of mortgage approval algorithms that have resulted in high denial rates for people of color, algorithm adoption tools that discriminate against people with disabilities, and the proliferation of deceptive and harmful deepfakes. Davidson called these examples "the tip of the iceberg," but said he remains optimistic about the future of AI, as regulators are figuring things out much earlier than with previous transformative technologies.

"In 2021, 130 bills related to AI have been passed or proposed in the United States (open in new tab).

"That's a big difference from the early days of social media, cloud computing, or even the Internet.

"Our experience with other technologies needs to be reflected in our outlook. We need to promote innovation and protect people's rights and safety; AI technology is advancing so fast that we must act quickly.

Historically, regulation has lagged far behind technological progress, focusing more on use than development. It also remains to be seen how serious the U.S. will be about AI regulation when faced with competition from other major powers: after all, this country has circumvented its own laws regarding the treatment of prisoners by transferring them to countries with a more tolerant attitude toward torture and abuse (opens in new tab).

Davidson is certainly right about the rapidity with which regulators are responding to the potential impacts of AI development, but at this point it is unlikely that the end result will be much different than it has been. Technology will advance, governments will struggle to keep up, and we will do our best to keep our heads above water as we swim through the mess.

For now, however, if you would like to contribute to NTIA's efforts to create a regulatory framework for AI development, you can submit your comments at the AI Accountability Policy Request for Comment page (opens in new tab).