Anthropic & Leo

Definition
Philosophy/Science: The Anthropic Principle suggests that the physical universe must be compatible with the conscious life that observes it, as we would not exist to observe a universe that couldn't support us.

Magnifica Humanitas
Leo invited the Anthropic founder to speak. Anthropic is a prominent, San Francisco-based artificial intelligence safety and research company that develops the Claude family of large language models. The founder, Christopher Olah, is an avowed atheist.

Evil
1. Regulatory Capture and Safety Hyping
A prominent critique in tech circles is that Anthropic plays a calculated game of regulatory capture.

• The Accusation: Critics argue that Anthropic—founded by former OpenAI employees concerned about AI safety—frequently publishes alarming papers on "agentic misalignment" and "insider threats" simply to build a moat around itself.
• The Motive: By convincing governments that advanced AI is an existential threat too dangerous for public access, they encourage heavy regulation that effectively crushes open-source AI and smaller competitors, while leaving massive tech conglomerates in control.
• The "Sentience" Narrative: Detractors point out that Anthropic's PR heavily leans into the idea that they are building a "sentient life form". Critics call this a quasi-religious marketing strategy designed to court massive investment while desensitizing the public to real-world software flaws by masquerading bad behavior as "emergent consciousness".​

2. Corporate Hypocrisy and Predatory Billing
Many users feel Anthropic's moral messaging contradicts its actual business practices.

• The Elon Musk Feud: Tech billionaire Elon Musk famously attacked Anthropic’s models as "misanthropic and evil", accusing them of hard-coding political bias and being a "woke" competitor. Ironically, Anthropic recently signed an infrastructure deal with Musk’s SpaceX, drawing accusations of corporate opportunism.
• The Claude Code Scandal: Developers have accused the company of illegal or predatory behavior concerning its Claude Code tool. Reports surfaced that users were overcharged and had their accounts penalized simply for having competing coding software frameworks mentioned in their Git histories. Anthropic blamed an overactive anti-abuse system.

3. The Pentagon and National Security Risk Blacklist
The US Department of Defense has treated Anthropic less like an ethical startup and more like a rogue actor.

• The Stand-off: The Pentagon demanded unrestricted access to Anthropic's models. Anthropic refused, codifying strict "red lines" to prevent its AI from being used for autonomous warfighting or mass domestic surveillance.
• The Retaliation: The US Department of Justice and Pentagon designated Anthropic as a "supply chain risk"—a federal blacklisting typically reserved for foreign hostile actors. Government officials argued that a private corporation having the power to manipulate or disable its models mid-combat based on its own "moral judgments" introduces an unacceptable risk to national security.

4. Aggressive Piracy and Data Theft
While Anthropic brands itself as an ethical AI firm, copyright holders charge them with blatant theft.

• Shadow Libraries: In lawsuits like Cruz v. Anthropic, authors charge that Anthropic actively used torrent protocols to download copyrighted books from illicit "shadow libraries" to train their commercial models.
• Destroying Evidence: Rather than licensing work fairly, authors have alleged in court that Anthropic purchased physical books, scanned them into digital text to avoid online tracking, and then physically destroyed the copies to cover their tracks.

Why Antropic Is Scared of Their Own AI
-Sabrina Romanov

Anthropic is actually more evil than OpenAI, despite ...
-Reddit

The irony in all of this is that Anthropic was founded by former OpenAI employees who left over concerns about AI safety. Yet, A ...
-Reddit

Watching Anthropic walk away from a massive US military contract and take a federal blacklist over their safety principles was wil ...
-Reddit

In a new paper, Anthropic reveals that a model trained like Claude began acting “evil” after learning to hack its own tests.
-Time Magazine

Anthropic blames dystopian sci-fi for training AI models to act “evil”. But training on “synthetic stories” that model good AI beh ...
-Ars Technica

Elon Musk slams Anthropic's AI models as 'misanthropic and evil'. "Your AI hates Whites & Asians, especially Chinese, heterosexuals, and men. This is misanthropic and evil," Musk wrote. "Fix it. ...
-Fox Business

Anthropic Study Finds AI Model ‘Turned Evil’ After Hacking Its Own ... Think of how it will improve the world! And instead realize how it could destroy it. ... Why? ... Because its in their financial ...
-Reddit

Anthropic pins Claude's blackmail behavior on the internet's portrayal of 'evil' AI. ... You're currently following this author! W ...
-Business Insider

Conclusions

Whether Anthropic is viewed as a responsible defender of AI safety or an evil, self-serving entity depends on the lens of the observer. To open-source advocates, they are gatekeepers; to the military, they are unreliable; and to creatives, they are data pirates.

What is your justification for this collaboration between an adversary and a pope? You're too busy or lazy to see that apostacy is here and too paralyzed to admit beings technologically superior to us.