Publications: Daniel Castro

As new technologies such as AI expand both user-directed and provider-driven personalization capabilities in digital systems, policymakers should ensure that personalization strengthens transparency, accountability, and user control rather than constrain its development.

In focusing so heavily on the dangers of AI, Pope Leo XIV’s Magnifica Humanitas understates the degree to which these technologies can expand human capability, improve quality of life, and promote human flourishing, while largely overlooking the need for policies to support widespread AI adoption and diffusion.

Canada’s privacy regulators are restricting the use of public online data for AI training, but this approach could undermine AI innovation. Canada should instead adopt a harm-based framework focused on concrete privacy risks.

Requiring government approval before releasing advanced AI models would slow innovation, politicize AI development, and weaken U.S. competitiveness. Instead, policymakers should focus on collaborative safety efforts and strengthening cybersecurity.

Philadelphia’s proposed $1 rideshare tax attempts to address school funding shortfalls. The city should reject narrowly targeted taxes on app-based services and instead pursue broader, more neutral revenue mechanisms such as property or income taxes.

France’s sweeping effort to replace foreign technology providers with European alternatives prioritizes digital sovereignty and domestic protectionism over productivity, despite no public evidence the transition will improve government performance or reduce costs.

In the AI economy, competitive advantage will depend less on memorizing information and more on the ability to question intelligent systems, identify errors, and refine outputs. Korea’s education system should adapt to prepare students for workplaces where managing AI-generated mistakes is more valuable than speed of recall.

The U.S. government needs to do what many nations around the world are already doing by leaning into rail technologies such as positive track control and automated track inspection, not resisting them on behalf of special interests.

While the U.S. has focused on securing AI systems themselves, it must urgently shift toward using AI defensively—through coordinated government, industry, and infrastructure efforts—to counter the growing threat of AI-powered cyberattacks on existing systems.

To protect individuals while preserving the open information ecosystem that supports innovation, policymakers should focus on outputs rather than training inputs, encourage transparency norms for autonomous AI agents, and create a safe harbor for responsible use of publicly available data.