WASHINGTON DC – OCTOBER 30: US President Joe Biden hands Vice President Kamala Harris a pen. … [+]
Artificial intelligence is gaining public policy attention. In October 2023, the Biden administration issued a presidential executive order on AI, directing federal agencies to work together to protect the public from potential AI-related harms. President Biden said in his State of the Union address in March 2024 that government enforcement officials would crack down on the use of AI to facilitate illegal price fixing. Congress is currently in the preliminary stages of considering legislation that could pave the way for future AI regulation.
The European Union “got even better” by adopting far-reaching AI legislation in March 2024. The law prohibits AI systems that are “considered to pose an unacceptable risk to the health, safety or fundamental rights of individuals.” The law also regulates general-purpose AI models that can be applied to a variety of tasks and imposes obligations depending on the level of risk they pose, or are believed to pose, to the public.
Should the federal government follow Europe's lead and introduce comprehensive AI regulation to prevent malicious activity? Simply put, not right now.
Existing laws prevent AI from being misused
First, keep in mind that, like all technologies, AI is already fully subject to generally applicable U.S. criminal and civil law.
In fact, the AI Executive Order of 2023 addresses these laws (civil rights, national security, antitrust, privacy, labor rights, health care, and foreign affairs) to combat any potential AI abuses. Emphasis is placed on the application of
This is more than rhetoric. For example, the Justice Department is reportedly “escalating” an antitrust investigation into a criminal scheme to use AI algorithm software to inflate and fix rental housing prices. And in January 2024, the Federal Trade Commission announced it was investigating the competitive effects of AI-related partnerships and investments.
These and similar government initiatives send a strong signal to AI developers to avoid harmful practices.
Regulation is the enemy of innovation and economic growth
Case-by-case enforcement of general laws allows us to target and address specific business misconduct without disrupting business plans and initiatives.
Regulation, in contrast, establishes a framework of rules that govern private sector behavior. This often translates into an inflexible, one-size-fits-all approach that ignores specific commercial circumstances and may not be able to respond in a timely manner to changes in technology and the business environment. Established business interests often benefit by leveraging their influence and manipulating this regulation to maintain their advantage.
A statistical analysis published in the Journal of Regulatory Economics in 2017 found that there were fewer new companies in highly regulated industries and that employment growth was slowing across all companies. Also,”[l]Large companies may even be successful in lobbying government authorities to tighten regulations to raise costs for smaller competitors. ”
As a 2023 American Economy Review article notes, regulation also tends to reduce innovation. As explained in a 2017 European Central Bank paper, innovation is the key to increased productivity and economic growth.
These factors, like the commercialization of the Internet in the past, are now accelerating innovation across a wide range of applications, leading to a rush to regulate new technologies that may require new management approaches. is strongly opposed. Specific cases highlight the benefits of avoiding too much regulation or too early regulation.
America's Internet Freedom vs. the EU's Repressive Approach
The US decision not to overregulate the internet in its early days is a major public policy success and should be studied carefully by public policy makers considering AI. Economic analyst Mohamed Muti considers the consequences of this decision in EconLog:
“In the mid-1990s, the Clinton administration made a wise choice. They believed that the Internet was an unregulated, “market-driven realm” with limited government involvement, a predictable, minimalist, and consistent environment. It declared itself limited to supporting and enforcing a simple legal environment. This policy has enabled a new generation of creative minds to explore the frontiers of business and commerce. This approach led to the success of the Internet and led to a surge of innovation. Today, the United States is home to some of the most innovative technology companies and has a vibrant Internet-based enterprise that provides countless benefits to consumers and small businesses. ”
By contrast, the European Union, which supports the precautionary principle of avoiding risks by introducing regulations early, has a poor innovation record.
Take, for example, the EU's efforts to regulate data privacy through the 2018 General Data Protection Regulation. GDPR requires companies to guarantee users' rights to access, consent, erasure, and data portability. There is already substantial evidence that they tend to impose high compliance costs, entrench incumbents, raise barriers to entry, and harm start-ups, new entrants, and small businesses.
This history should give pause to anyone who believes that an AI law could help position Europe as an AI leader.
Is now really the time?
Regulation is justified if it is proven that its benefits outweigh its costs. Of course, cost-effectiveness evaluations are incomplete and prone to error. In his famous 1969 paper, renowned economist Harold Demsetz compares idealized regulatory proposals (assuming they work perfectly) with the actual outcomes of the current unregulated system. I warned you against it.
A history of harmful regulatory deficiencies underscores the wisdom of Mr. Demsetz's warning. If the US implements her AI regulations, it will likely slow down innovation in technology systems that are supposed to provide significant benefits to society. Even a small decline in the growth rate of AI could lead to huge social welfare losses in the long run. Imagine how much of the benefits of innovation would have been lost if governments had decided to effectively control the Internet 30 years ago.
What about AI-related harms? That's highly speculative at this point. Additionally, the federal government is closely monitoring the AI sector and stands ready to apply existing targeted legal sanctions at the first sign of trouble.
In summary, today's AI landscape is characterized by rapidly increasing benefits and uncertain costs, which are likely to be adequately addressed under existing law should problems arise.
There may (or may not be) a time when new, unexpected and serious AI-related issues arise that are best addressed through targeted regulatory solutions. However, there is currently no basis for regulating AI today.
