Dean Ball, a member of the board of directors, The Alexander Hamilton Institute for the Study of Western Civilization (AHI), points to a way in which artificial intelligence (AI) can be intelligently regulated. But Mr. Ball warns in the second essay in his two-part series on artificial intelligence’s near-term future that AI should not be micromanaged nor unduly exposed to potentially devastating lawsuits.

Since early 2024, Mr. Ball—an AHI undergraduate fellow when he was at Hamilton College, and thus an “AHI alum”—has written Hyperdimensional, a highly successful online newsletter about artificial intelligence. He is a research fellow specializing in AI at the Mercatus Center at George Mason University.

Artificial intelligence, he writes, is comparable to the “major foundational technologies” on which modern life is built, such as railroads, electricity, and the internet. It should receive something like the legal protection these industries enjoy due to the fact that public policy has “eliminated or severely limited” their exposure to tort lawsuits for people’s misuse of their products.

If a high-tech company is testing a new AI model and, so to speak, it escapes from the laboratory and starts to “scam people on the internet,” the company should face liability. “But if models from, say, Anthropic are fundamental parts of economic output in virtually every industry, it is unreasonable to force Anthropic to bear liability risk for all this activity by their customers.”

Without a wisely crafted framework for its regulation, Mr. Ball sees the possibility of extremely “complicated and expensive litigation” when AI is alleged to cause harm. In addition, the nature of tort litigation “does not cleanly channel lawsuits toward the most responsible entity.”

A further cause for concern is that we have seen the rise of litigation finance, “investors who fund tort lawsuits in exchange for a share in the damages awarded.” This allows too much room for abuse: “Think … about the clever things an adversarial government (or government-linked private investors) could do to hobble the American AI industry with this tool at their disposal.” And yet another danger: “the shape of tort law can change overnight with the political winds.”

Mr. Ball agrees that the government should provide incentives for firms to build advanced AI responsibly. But he thinks it would be best to try to avoid the problems associated with centralized regulation.

He has proposed “a private governance framework for AI” whereby “companies would opt in to receiving certifications” from private entities the government oversees. “Ideally, there would be multiple such bodies competing with one another … [and much] flexibility in how certifications could be designed and enforced.” Tort liability adjudicated in the courts would remain as a backup means of regulation.

Mr. Ball concludes by suggesting that AI in free societies can have a substantial advantage over AI in unfree societies simply by being governed more reasonably—and this can happen if free countries like the United States avoid “micro-managing” it. But because artificial intelligence will be so formidably capable, he adds, micromanagement of it will become increasingly “tempting and feasible.”