Request for Information for a new US AI Action Plan (OSTP RFI)

Link post

TL;DR: The Office of Science and Technology Policy is seeking public input (deadline: March 15, 2025) for a new AI Action Plan following Trump’s recent AI executive order. While the new administration’s focus on reducing oversight may limit receptiveness to safety-focused input, the low effort required to submit could make it worth trying, especially if safety considerations can be framed in alignment with the administration’s priorities of maintaining US AI leadership and competitiveness.

The Office of Science and Technology Policy (OSTP) is seeking public input on the Development of an AI Action Plan following the recent Executive Order on Removing Barriers to American Leadership in Artificial Intelligence (EO 14179). This RFI presents an opportunity for organizations and individuals to provide input on the development of the so-called AI Action Plan that the new EO mandates. However, the framing differs notably from previous AI-related executive orders (see below for context).

The RFI is seeking input on a wide range of AI policy topics, including:

Hardware and chips
Data centers and energy consumption
Model development and open source
Explainability and assurance
Cybersecurity and privacy
Risk management and governance
Technical and safety standardsopen-source
National security applications
Research and development priorities
Education and workforce development
International collaboration
Export controls

Deadline: March 15, 2025 at 11:59 PM ET

Some context

This RFI follows the January 2025 revocation of the previous Biden-Harris AI Executive Order (EO 14110) on Safe, Secure, and Trustworthy Development of AI. The new administration’s stated focus is on “sustaining and enhancing America’s AI dominance” while reducing “unnecessarily burdensome requirements” on private sector AI development.

The AI Action Plan, mandated by the EO, must be developed within 180 days (by July 2025) and will involve coordination across multiple agencies including the Office of Science and Technology Policy, the White House AI and Crypto office, and the National Security Council. The Plan is meant to operationalize the administration’s AI policy goals of promoting “human flourishing, economic competitiveness, and national security” while removing what the administration views as barriers to AI innovation.

While the emphasis on reducing oversight and accelerating development may limit receptiveness to certain safety-focused recommendations, there might still be value in providing careful, well-reasoned input on risk management, safety standards, and governance frameworks. I’d approach this as mostly hits-based, considering that the effort required is pretty low.

If you’re drafting a submission, consider whether you might be able to frame key safety considerations in ways that would resonate with the current administration’s priorities. You might find this blogpost on AI policy considerations under a Trump admin to be helpful.

Practical details

Submissions can be sent to ostp-ai-rfi@nitrd.gov (include “AI Action Plan” in the subject line).

Note that there are some pretty strict formatting requirements:

15-page maximum.
12-point or larger font.
Include page numbers.
One submission per individual/organization.
Must include a statement approving public dissemination:
- “This document is approved for public dissemination. The document contains no business-proprietary or confidential information. Document contents may be reused by the government in developing the AI Action Plan and associated documents without attribution”
Avoid including any confidential information.