Anthropic AI Policy Recommendations to White House - Latest News

Anthropic Submits National AI Policy Recommendations to the White House

Following the discreet removal of previously pledged AI policy commitments from its website, Anthropic has presented recommendations to the White House. These suggestions aim to position the United States to maximize the economic advantages offered by advancements in artificial intelligence.

Key Recommendations for National AI Policy

The proposals put forth by Anthropic encompass several crucial areas. These include the continued operation of the AI Safety Institute, initially created during the Biden administration.

Furthermore, the company advocates for the National Institute of Standards and Technology (NIST) to spearhead the development of national security assessments specifically designed for highly capable AI models.

A dedicated governmental team should also be established, according to Anthropic, to proactively identify and analyze potential security weaknesses inherent in AI systems.

Focus on Security and Infrastructure

Strengthened controls on the export of AI chips are also a key component of Anthropic’s recommendations. Specifically, restrictions on the sale of Nvidia H20 chips to China are proposed, citing national security concerns.

To adequately support the energy demands of expanding AI data centers, Anthropic suggests the U.S. set a national goal of constructing an additional 50 gigawatts of power capacity dedicated to the AI sector by the year 2027.

Alignment with Previous Policy and Subsequent Changes

Many of these policy suggestions demonstrate a strong correlation with the AI executive order issued by former President Biden. This order was rescinded by the Trump administration in January.

Those critical of the Biden order, and aligned with the Trump administration, voiced concerns that its reporting stipulations were excessively burdensome.

The current recommendations from Anthropic represent a continued effort to shape the national conversation surrounding AI regulation and development.