TABLE OF CONTENTS I. INTRODUCTION II. THE TROUBLE WITH AI A. What Is AI? B. The Problematic Characteristics of AI 1. Autonomy, Foreseeability, and Causation 2. Control 3. Research and Development: Discreet, Diffuse, Discrete, and opaque C. A Role for the Law? III. INSTITUTIONAL COMPETENCE A. Legislatures 1. Democratic Legitimacy 2. Lack of Expertise 3. Delegation and oversight B. Agencies 1. Flexibility 2. Specialization and Expertise 3. Independence and Alienation 4. Ex Ante Action C. The Common Law Tort System 1. Fact-Finding 2. Reactive (and Reactionary) 3. Incrementalism 4. Misaligned Incentives IV. A REGULATORY PROPOSAL A. The Artificial Intelligence Development Act B. The Agency C. The Courts' Role V. CONCLUSION I. INTRODUCTION It may not always be obvious, but we are living in the age of intelligent machines. Artificial intelligence (AI) permeates our lives in numerous subtle and not-so-subtle ways, performing tasks that, until quite recently, could only be performed by a human with specialized knowledge, expensive training, or a government-issued license. Driverless cars have been approved for road operation in four states and the District of Columbia; (1) their inevitable arrival on the consumer market may revolutionize road transportation. Autonomous machines can execute complex financial transactions, flag potential terrorists using facial recognition software, and (most alarmingly for this author and his legal contemporaries) perform document review. (2) More mundanely, computer chess engines can defeat the strongest human players in the world, and Google Translate can generate passable English translations of Le Monde articles. In fact, robot journalists may even have written the Le Monde articles themselves. (3) The increasing ubiquity and rapidly expanding commercial potential of AI has spurred massive private sector investment in AI projects. Firms such as Google, Facebook, Amazon and Baidu have got into an AI arms race, poaching researchers, setting up laboratories and buying start-ups. (4) With each passing month, AI gains footholds in new industries and becomes more enmeshed in our day-to-day lives, and that trend seems likely to continue for the foreseeable future. (5) The potential for further rapid advances in AI technology has prompted expressions of alarm from many quarters, including some calls for government regulation of AI development and restrictions on AI operation. (6) That in and of itself is hardly surprising; fear of technological change and calls for the government to regulate new technologies are not new phenomena. What is striking about AI, however, is that leaders of the tech industry are voicing many of the concerns. Some of the concerns stem from the familiar fears of technological unemployment (7) and the potential for new technologies to be misused by humans. (8) But many of the fears cut much deeper. In an interview at MIT's 2014 AeroAstro Centennial Symposium, Elon Musk eschewed the skepticism of regulation that characterizes most of Silicon Valley's business titans and suggested that some government intervention might be wise in the case of artificial intelligence: I think we should be very careful about artificial intelligence. If I had to guess at what our biggest existential threat is, it's probably that .... I'm increasingly inclined to think there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don't do something very foolish. (9) Other prominent figures in the tech world--most notably Bill Gates and Steve Wozniak--have voiced similar concerns regarding the long-term risks of AI. (10) At the very least, more mundane legal issues surrounding AI seem likely to crop up in the near future. Who (or what) will be held liable when an autonomous vehicle causes an accident? …