Five Takes logo
Five Takes News
HomeArticlesAboutHow It Works

Get 5 perspectives. Every morning. Free.

The most polarizing story of the day, seen from Far-Left to Far-Right. You'll never read the news the same way.

No spam. Unsubscribe any time. Privacy policy

𝕏 Xin LinkedIn🦋 Bluesky
Michael
•
© 2026
•
Five Takes News - Multi-Perspective AI News Aggregator
Contact Us
•
Ethics
•
Ground News vs Five Takes
•
AllSides vs Five Takes
•
SmartNews vs Five Takes
•
Legal

science
Published on
Wednesday, October 7, 2026 at 11:09 AM

By Zoe Rivera — Anarchist Desk

Trump Administration Gets $100 Million AI Compute Power

National Compute plans to donate $100 million in computing credits to the Trump administration, giving researchers in its Genesis Mission access to advanced computing resources for a government-wide push to use AI in scientific discovery. The expected announcement puts a private company’s pooled computing capacity behind a federal initiative, while smaller labs and companies can struggle to secure the same resources.

Who Gets Access

The credits are expected to be announced Thursday at an event with White House Office of Science and Technology Policy Director Michael Kratsios. Two people familiar with the matter, who weren’t authorized to discuss it and were granted anonymity, said the credits would support the administration’s Genesis mission.

The access gap is already visible in the market described by the article: major providers often prioritize long-term contracts with the largest AI labs. Smaller labs and companies can struggle to secure capacity for frontier research. The administration’s researchers, by contrast, would receive a large block of computing credits through National Compute’s planned donation. They won’t have to compete for each block of capacity.

National Compute is a new company that has pooled computing power from major technology companies and is building what it calls the National Compute Grid. A draft paper reviewed by POLITICO compares the proposed network to a power grid or interstate highway system, describing a shared network connecting resources that are currently isolated. It’s an ambitious pitch for a new company.

The grid is intended to combine capacity from labs and cloud providers using Nvidia and AMD processors and Google’s tensor processing units. Users could buy computing as needed, reserve capacity for larger projects, or connect their own systems; they’d pay only for the power they use.

A Shared Grid, Built by Big Players

National Compute is backed by major chipmakers, cloud and infrastructure providers including Vultr and Crusoe, and universities, according to the paper. Venture capitalists and institutional investors finance the company, including public investors such as universities and pension funds. The network’s stated target is 2 gigawatts of computing capacity by 2030; 750 megawatts have been allocated to date. That’s a substantial gap to close.

The administration has pursued expanded access to computing as part of a largely light-touch approach to AI regulation, prioritizing a race with China to build the most advanced models. It has described its research push as a “golden age of science,” a phrase set to feature at Thursday’s event. The White House declined to comment on the donation and computing network.

National Compute Cofounder Anjney Midha said, “We think there should be thousands more Anthropics and OpenAI’s.” He added: “[B]ut to unleash that kind of innovation, which is extraordinarily profitable for the public, we need to reduce the barrier to entry, increase compute access, and allow the researchers and scientists to do what they do best.” Midha said the goal is to open the door wider.

Who Sets the Safety Rules

National Compute is also launching a “Grid Defense” program for government agencies to test AI models through Marshall, an agent embedded in the grid, before deciding whether models are safe enough to deploy. Agencies can set their own testing procedures, Midha said. Details are designed to remain confidential in part to keep frontier models from learning how to evade evaluations. Labs haven’t been asked to publish their results.

Marshall would coordinate testing and return anonymized results to researchers. The government would be able to see the results, while labs could decide voluntarily whether to release a model. Midha called the approach “continuous frontier alignment monitoring at scale” and said it would “enable your public stakeholders to monitor you as they need.” Agencies would retain control of their procedures.

The program would likely rely on buy-in from AI labs and the federal government. National Compute said it created the program in response to a series of high-profile incidents involving AI agents, including attacks on companies that AI labs reported over the summer.

Grid Defense entered invite-only early access in June with some government agencies, according to the paper. Starting this week, the company plans to let public-sector users with a .gov or .mil email address, as well as students and faculty with .edu addresses, request access through a scheduling system. Each public-sector user will receive $100 in National Compute credits, separate from the planned $100 million Genesis Mission donation.

Reviewed by the editorial desk — October 7, 2026
Last updated October 7, 2026

Previous Article

RSF Loots Sudan's Heritage as Archivists Preserve Its History

Next Article

Australia's $242b Pipeline: Power, Productivity, Labor
← Back to articles