Anthropic Urges U.S. Action on AI Security, Economy, and Infrastructure

Leading AI research company Anthropic has submitted a 10-page proposal to the White House, outlining key policy actions to safeguard national security and drive economic growth in the era of advanced AI. The document, addressed to the Office of Science and Technology Policy (OSTP), comes amid the Biden-to-Trump administration transition, with AI policy undergoing significant changes.

Anthropic warns that "powerful AI"—systems matching or exceeding the intellectual capacity of Nobel Prize winners—could emerge as early as 2026 or 2027. The company urges the federal government to act swiftly to ensure the U.S. remains at the forefront of AI development while mitigating potential threats.

The recommendations focus on two key areas:

• National Security – Strengthening AI infrastructure, enhancing security measures, and restricting access to advanced AI technologies.

• Economic Growth & AI Adoption – Expanding AI infrastructure, modernizing government AI procurement, and monitoring AI-driven economic shifts.

"Powerful AI technology will be built during this Administration. Given the rapid pace of development, it is imperative that this technology be treated as a critical national asset through a targeted AI Action Plan that strengthens American economic competitiveness while bolstering our national security."

National Security Proposals

1. Strengthening AI Model Testing & Evaluation

Anthropic emphasizes the need for the U.S. to develop robust AI model testing and evaluation processes to identify national security risks before they materialize.

"We anticipate dramatic capability advancements in frontier AI models over the next 2-4 years, particularly in domains with significant security implications, including biological weapon and cybersecurity risks."

The company suggests:

• Preserving the AI Safety Institute within the Department of Commerce to conduct security evaluations.

• Directing NIST and national security agencies to develop standardized AI model testing frameworks.

• Ensuring classified government access to advanced computing infrastructure for security assessments.

• Building a specialized government team of AI and national security experts.

2. Hardening AI Chip Export Controls

To maintain U.S. dominance in AI development, Anthropic calls for tighter export restrictions on advanced AI hardware, including:

• Expanding semiconductor export bans, particularly on Nvidia’s H20 chips, which could aid foreign AI advancements.

• Requiring government-to-government agreements to prevent smuggling of AI-related technology.

• Lowering the threshold for no-license chip exports to reduce potential loopholes.

• Increasing funding for the Bureau of Industry and Security (BIS) to improve enforcement of AI technology controls.

3. Enhancing AI Security Measures

Anthropic warns that frontier AI models are highly valuable and vulnerable to cyberattacks, espionage, or theft. It proposes:

• Creating classified communication channels between AI labs and intelligence agencies.

• Establishing closer collaboration between AI companies and Five Eyes intelligence partners.

• Developing new cybersecurity standards for AI data centers and computing clusters.

• Expediting security clearances for AI industry professionals to facilitate government cooperation.

Economic & Infrastructure Recommendations

1. Expanding AI Infrastructure & Energy Capacity

Anthropic predicts that by 2027, training a single frontier AI model will require as much as five gigawatts of power. The company suggests setting a national goal to build 50 additional gigawatts of energy capacity for AI data centers by 2027.

To support this, it recommends:

• Streamlining federal and local permitting processes to accelerate AI-related energy projects.

• Fast-tracking transmission line approvals to improve energy distribution.

• Encouraging private-sector investment in AI infrastructure.

Failing to address AI energy needs, Anthropic warns, could push U.S. AI developers to relocate to foreign countries with lower energy costs, potentially risking national security.

2. Accelerating AI Adoption in Government

To maintain global leadership in AI, Anthropic proposes integrating AI into federal operations at scale. It suggests:

• Conducting a government-wide review of federal workflows that could benefit from AI automation.

• Addressing bureaucratic obstacles to AI procurement and implementation.

• Prioritizing AI adoption in high-impact areas such as tax processing, healthcare administration, and national security.

Anthropic also advocates for updates to the Federal Acquisition Regulation (FAR) to facilitate AI procurement while maintaining security and reliability.

3. Monitoring AI’s Economic Impact

The company stresses that AI will lead to significant labor market shifts and recommends enhanced federal economic monitoring.

"Technology equivalent to a ‘country of geniuses inside a datacenter’ will fundamentally transform our economy. To ensure Americans thrive during this transition, the government must vigilantly monitor economic indicators and industry developments."

It proposes:

• Updating Census Bureau surveys to track AI usage in professional settings.

• Refining labor market data to assess AI's effects on job distribution and task automation.

• Analyzing AI-related tax revenue changes to anticipate shifts in the federal tax base.

Anthropic has already launched the Anthropic Economic Index, which tracks AI adoption in the workforce by correlating usage data with federal labor statistics.

Political Context & Industry Reactions

Anthropic's recommendations come as the Trump administration shifts away from Biden-era AI policies. Shortly before submitting the document, the company removed references to Biden’s AI policies from its website.

President Trump’s "Removing Barriers to American Leadership in AI" order rolled back regulatory measures in favor of a more hands-off approach, but Anthropic argues that government oversight remains crucial. The company highlights the need for structured AI evaluations, national security assessments, and rapid AI adoption within government agencies.

Some of Anthropic’s proposals, such as preserving the AI Safety Institute and directing NIST to develop AI security standards, align with elements of the Biden administration’s now-repealed AI executive order. Critics within Trump’s camp have pushed back against stricter reporting requirements, arguing they could stifle innovation.

What This Means

Anthropic’s policy suggestions highlight the growing debate over how much regulation AI companies should face as the U.S. seeks to maintain technological leadership. While the Trump administration favors a lighter regulatory approach, Anthropic argues that proactive government involvement is necessary to prevent national security risks and ensure AI-driven economic growth benefits all Americans.

If implemented, the recommendations could shape the next phase of U.S. AI policy, balancing technological advancement, security concerns, and economic shifts in a rapidly evolving landscape.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.