Here’s a recap of what US legislators accomplished on the tech (including AI) regulation front in 2018. https://www.thenewstribune.com/news/local/article223939600.html
Spoiler alert: there’s a lot of words in the article, but the answer is nothing. OK so what legislative excitements will abound in 2019? Spoiler alert: not much. This, according to the article, often because legislators don’t have much knowlege about the subject matter and are often held back by big lobbying (fair). Also bewildered accoring to the headline (also fair). But that doesn’t seem like an answer to much. There are lots of laws on the books that congressmen and members of parliament and senators and other legislators don’t know much about and didn’t know much about when those laws got passed. And this blog has demonstrated lots of thoughtful considerations on AI by various government institutions. So what gives? Why the intertia?
I might suggest that one impetus that is lacking is any kind of public crisis around AI. Why regulate AI if there don’t seem to be any problems that are getting consituents to call your office consistently or that shift large blocks of voters?
But there is also another view that seems to be often behind this. And that is the suggestion that premature regulation is deleterious to innovation and progress. Now no one wants things that are premature and deleterious, but is this really the case in terms of regulation of AI? Couldn’t we have some basic building blocks of regulation? Wouldn’t that add certainty rather than leaving an uncertain legal void in which companies are asked to operate? We see the problems start in this analysis with the hopeful intent to define AI terms, but that disappears quickly with “narrow” AI being defined by example rather than meaningfully. And there are a series of straw man arguments here; nobody is suggesting that a risk-taking free enterprise system should be curtailed with special penalization of the adoption of advanced technologies.
But there are some nice thoughts along the way. Referencing AI as a general purpose technology might help those bewildered legislators who so often do focus on individual companies and how they operate in their hearings. And it is true that regulatory restraint does not necessarily lead to harm for consumers.
This blog has noted US government reports that have called for legislative action that cuts across different jurisdictional levels to create an even framework for all players. It is interesting that his article notes the opposite as being the strategy. Certainly that would be consistent with past industry legislative approaches. But the question arises, in a technology that will likely operate pan-nationally and internationally, will that approach be sufficient to keep up with the legislative approaches of countries that are more jurisdictionally homogeneous.
The good news, if you’re a go slow kind of guy, is that the legislators have got stand still down pat.