Anthropic donates $20 million to the "Public Priority Action Organization" to promote AI defense

Anthropic announces a $20 million donation to Public First Action (Public Priority Action Organization, abbreviated PFA) to support initiatives related to artificial intelligence governance. Anthropic states that this funding will be used to promote public education on AI and establish legal frameworks, particularly for regulating potential risks such as preventing automated cyberattacks and the use of AI technology for manufacturing weapons and other AI risk scenarios.

Anthropic: AI development poses risks, and governments need to promote policies to ensure safety

Anthropic points out that AI has the potential to bring widespread benefits in fields such as scientific research, technological innovation, medical development, and economic growth, but its potential risks cannot be ignored. These risks stem from the misuse of technology, such as being used for automated cyberattacks or, in the future, for creating dangerous weapons. There is also the possibility that as models improve, they may deviate from user intent, leading to uncontrollable behaviors. As AI models continue to evolve rapidly, managing potential risks while promoting innovation will be a key issue in AI development.

AI model capabilities are advancing quickly, from chatbots to AI agents capable of executing multi-step, complex tasks. Anthropic states that to meet the challenges posed by continuously evolving AI models, the company is raising the bar for recruiting software engineers, providing new hires with high-difficulty tests to ensure they can handle the fast-changing AI landscape.

In light of this, Anthropic believes that the rapid proliferation of AI is surpassing most previous technologies. However, the pace of AI policy development remains slow. Currently, there is no comprehensive official regulatory mechanism, nor a federal-level overarching framework, resulting in fragmented policy discussions. Anthropic states that AI policies in the coming years will impact issues such as the labor market, child protection, national security, and international power dynamics. Therefore, it is necessary to establish flexible and responsive regulatory systems to balance innovation and risk.

Donating $20 million to Public First Action Organization, non-partisan

Anthropic states that the donation of $20 million to Public First Action is not motivated by political stance nor solely for the company’s own interests. It emphasizes that AI developers have a responsibility to ensure that technology aligns with public interests, and this donation reflects that stance.

Public First Action is not a traditional lobbying organization but a political advocacy group. Its focus areas include protecting the rights of children, workers, and the public; strengthening U.S. export controls on advanced semiconductors; increasing transparency in AI laboratories; and reducing significant AI risks. Public First Action strongly supports state and local legislative efforts to address emerging AI issues and opposes federal government actions that hinder state progress in AI without sufficient federal safeguards. The organization is composed of bipartisan American politicians. Notably, although Anthropic states that the donation is not intended to support any particular political party, Public First Action recently supported Tennessee Republican Senator Marsha Blackburn’s campaign for governor.

This article about Anthropic donating $20 million to Public First Action to promote AI defense first appeared on ABMedia, Chain News.