Publications

The United States could pour an estimated $10.3 trillion into the infrastructure needed to support artificial intelligence between 2025 and 2032.

Developing common AI safety standards does not, by itself, undermine competition. Policymakers should not force companies to choose between competing vigorously and working together to prevent serious safety risks. Antitrust law should allow both.

More troubling than any effects on competition the decree will have is what it does to antitrust itself: It uses competition law to enact industrial policy goals that have little to do with competition.

Japan was a competitor inside the U.S. alliance system. China is a rival trying to replace it. Waiting for China to repeat Japan’s slowdown is not a strategy.

Foreign online piracy causes significant economic harm to U.S. creative industries, yet American courts lack the authority to block foreign pirate sites. Congress should adopt carefully designed site-blocking legislation that targets large-scale piracy while minimizing the risk of overblocking.

Europe cannot match the United States and China in closed AI models, but it has an opportunity to lead in open models. The EU should remove regulatory and procurement barriers and foster a global open-model ecosystem.

The Proposed Amendment would have the effect of “protecting” startups by frustrating larger firms from acquiring their founders and employees. Adopting such a policy would predictably chill incentives to innovate.

Democrats should approach data centers as they have clean-energy infrastructure: take legitimate local concerns seriously and mitigate their impacts without allowing opposition to become a blanket rejection of development. As AI drives growing demand for data centers, policymakers should balance community protections with the economic, technological, and clean-energy benefits these investments can provide.

Mandating broader transparency around algorithmic or personalized pricing requires congressional legislation to weigh costs and benefits, and the FTC should revise its policy statement to target law violations without restricting legitimate personalized pricing.

Between 1996 and 2025, Bayh-Dole-enabled technology transfer from university and nonprofit research benefiting from federal funding contributed as much as $3.3 trillion to U.S. gross economic output, $1.7 trillion to gross domestic product, and 7.6 million job-years of employment.