It’s alarming to see the US Government taking a hands-off and behind-the-scenes approach to the regulation of artificial intelligence (AI) technology. AI is promising to be an enabling technology that not only replaces human judgement in a variety and increasing amount of professions and institutions. It will also be used to coordinate societies and its resources and even conduct warfare through invisible policies embedded in complex code. Hands-off or half-baked approaches will lead to a loss of technical expertise, moral leadership, and ultimately: power.
Over the last few months, China announced their plan for exponential growth in AI research and applications over the next decade, Russia equates AI prowess to global domination, Canada and France have developed coherent national strategies, and Europe takes on moral leadership in the development of AI. Leadership in AI technology is increasingly seen as a geopolitical necessity. A White House representative recently justified their hands-off approach, however, by stating that AI technology is still in its infancy and should therefore not be burdened with regulation. Their aim is mostly to identify regulatory burdens for the development of AI technologies in specific sectors, and abolish them.
This view is problematic for several reasons. Reducing regulation does not mean that AI development is supported directly. Regulation does not equate to “red tape” and can, in fact, be constructive for sectors and technology. For example, the Obama administration was the first national government to publish an AI strategy. Their policy offered a list of concrete, constructive and actionable recommendations for a coordinated approach in developing and meaningful advances in AI technology meaningfully. It appears that only a few of these recommendations are in the process of being implemented, mostly around funding for research and defenseuses.
Regulation and policy also set the moral compass of technology development. As I frequently argue, computer scientists who develop AI technologies should not be the ones deciding, consciously or not, which values should be embedded in the technologies that shape our world. By setting out a national strategy for the development of technology, technologists can be incentivized to operationalize certain values over others to benefit the public good.
Immigration is another huge regulatory issue that has become a hot potato in the US for other political reasons. In the global arms race that AI technology is becoming, it is important for countries to not only train their own developers but also to attract talent from abroad. This is becoming increasingly hard, if not impossible, for many experts.
To their partial credit, the White House’s focused and more practical approach to AI policy is useful to understand how a variety of regulations interact with the developing technical landscape. However, this is what smaller countries–who are recipients of AI technology–should be doing. Countries that want to maintain leadership, need to articulate strategy, communicate it widely, and then put their money where their mouth is. While some representatives in Congress are pushing, the White House appears to be failing at offering vision, strategy, and leadership.