Tag Archives: Claude Anthropic AI and violence

How NGOs are Used in the AI Propaganda Wars

The artificial-intelligence startup Anthropic is doubling its spending on the midterm elections in the United States to $40 million by contributing to Public First Action, a political group pushing for government-imposed safeguards on powerful AI models….Public First says it is a counterweight to the other major AI spender, Leading the Future, which is funded by tech titans including a top executive at OpenAI and venture firm Andreessen Horowitz. Leading the Future is more aligned with the White House’s views on AI, pushing for industry-friendly rules that would allow AI to develop with fewer barriers….

Critics of Anthropic have said the company wants to use regulations to hurt competitors. They argue that Amodei, the founder of Anthropic, is hypocritical for warning about AI’s dangers while continuing to develop better models. Amodei and the company say their focus on model safety is genuine.

The clashing political groups, through networks of super PACs**, have spent millions of dollars on advertising in dozens of U.S. primaries. ..Leading the Future, which is reusing parts of a political playbook deployed by a pro-crypto super PAC network tin 2024, has reported raising about $75 million but has announced millions more in pledges. Public First said at the end of June it had raised $80 million.

Excerpt from Laura J. Nelson,Anthropic Doubles Midterm Spending to $40 Million to Push AI Regulation, WSJ, July 22, 2026

++Independent expenditure-only political action committees, commonly known as super PACs, are a type of political action committees (PACs) in the United States. Unlike traditional PACs, super PACs are legally allowed to fundraise unlimited amounts of money from individuals or organizations for the purpose of campaign advertising (see wikipedia)

If You Play with Fire, You ‘Il Get Burnt: Lessons from Anthropic

Anthropic’s artificial-intelligence tool Claude was used in the U.S. military’s operation to capture former Venezuelan President Nicolás Maduro on January 3, 2026, highlighting how AI models are gaining traction in the Pentagon. The mission to capture Maduro and his wife included bombing several sites in Caracas in January 2026. Anthropic’s usage guidelines prohibit Claude from being used to facilitate violence, develop weapons or conduct surveillance. The deployment of Claude occurred through Anthropic’s partnership with data company Palantir, whose tools are commonly used by the Defense Department and federal law enforcement

Excerpt from Pentagon Used Anthropic’s Claude in Maduro Venezuela Raid, WSJ, Feb. 15, 2025

See also Trump orders government to stop using Anthropic in battle over AI use, WSJ Feb. 27, 2026