Saved in:
Bibliographic Details
Main Author: Hastings-Woodhouse, Sarah
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.21082
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908470062612480
author Hastings-Woodhouse, Sarah
author_facet Hastings-Woodhouse, Sarah
contents Recent AI progress has outpaced expectations, with some experts now predicting AI that matches or exceeds human capabilities in all cognitive areas (AGI) could emerge this decade, potentially posing grave national and global security threats. AI development is currently occurring primarily in the private sector with minimal oversight. This report analyzes a scenario where the US government centralizes AGI development under its direct control, and identifies four high-level priorities and seven safety features to reduce risks.
format Preprint
id arxiv_https___arxiv_org_abs_2507_21082
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Safety Features for a Centralised AGI Project
Hastings-Woodhouse, Sarah
Computers and Society
Recent AI progress has outpaced expectations, with some experts now predicting AI that matches or exceeds human capabilities in all cognitive areas (AGI) could emerge this decade, potentially posing grave national and global security threats. AI development is currently occurring primarily in the private sector with minimal oversight. This report analyzes a scenario where the US government centralizes AGI development under its direct control, and identifies four high-level priorities and seven safety features to reduce risks.
title Safety Features for a Centralised AGI Project
topic Computers and Society
url https://arxiv.org/abs/2507.21082