Skip to content

Questions about Claude (language model)

Short answers, pulled from the story.

What is Claude and when was it first released?

Claude is an AI assistant developed by Anthropic, first released to selected users in March 2023. Claude 2, released in July 2023, was the first version available to the general public. Since Claude 3 in March 2024, each generation comes in three sizes: Haiku, Sonnet, and Opus, from smallest to most capable.

What is constitutional AI and how does Anthropic use it to train Claude?

Constitutional AI is a training technique Anthropic uses to guide Claude's behavior through a written document of principles rather than relying extensively on human feedback. Anthropic's first AI constitution was published in 2022; the 2023 update contained 75 guidelines partly drawing on the 1948 UN Universal Declaration of Human Rights. The 2026 edition expanded to 23,000 words.

What is Claude Code and how successful has it been?

Claude Code is an agentic command-line tool that lets developers delegate coding tasks from their terminal using natural language prompts. It launched as a preview in February 2025 and became generally available in May 2025 alongside Claude 4. By July 2025, Anthropic reported a 5.5 times increase in Claude Code revenue from its launch.

Why did the U.S. Department of Defense declare Anthropic a supply chain risk?

Defense Secretary Pete Hegseth declared Anthropic a supply chain risk on the 27th of February 2026, after Anthropic refused to remove prohibitions on using Claude for mass domestic surveillance and fully-autonomous weapons. President Trump directed all federal agencies to stop using Anthropic technology within six months. On the 26th of March 2026, federal judge Rita F. Lin issued a temporary injunction, stating the Pentagon's actions appeared to be classic First Amendment retaliation.

What is Claude Mythos and what did Project Glasswing accomplish?

Claude Mythos is an advanced Anthropic model released in limited preview on the 7th of April 2026 through Project Glasswing, made available to 11 companies for cybersecurity vulnerability research. Within two weeks, Mozilla announced it had found and patched 271 security vulnerabilities in Firefox using Mythos Preview. The NSA also used Mythos despite the Department of Defense having blacklisted Anthropic.

What did the Claude Opus 4.6 C compiler experiment demonstrate?

In February 2026, Anthropic researcher Nicholas Carlini reported that 16 Claude Opus 4.6 agents working together had written a C compiler in Rust from scratch, capable of compiling the Linux kernel. The experiment cost nearly $20,000. Carlini noted the compiler was not very efficient but identified Opus 4.6 as the first model capable of completing the task.

Up next

Queue