AnswersPublished2 months agoLast edited 2 months ago26 sources

How Elon Musk Went from Calling Anthropic ‘Evil’ to Powering Its AI — and What It Means for the Industry

Elon Musk publicly attacked Anthropic as 'misanthropic and evil' in February 2026, but by May 2026, SpaceX leased the company the entire Colossus 1 supercomputer in a deal reportedly worth $4 billion. Three key developments enabled the shift: (1) Anthropic's $1.2 trillion implied valuation on secondary markets by Ma...

Search & fact-check with Studio Global AI Browse more Trending pages

Research answer

Key takeaways

Elon Musk publicly attacked Anthropic as 'misanthropic and evil' in February 2026, but by May 2026, SpaceX leased the company the entire Colossus 1 supercomputer in a deal reportedly worth $4 billion.
Three key developments enabled the shift: (1) Anthropic's $1.2 trillion implied valuation on secondary markets by May 2026, surpassing OpenAI; (2) the launch of the Mythos class models (Claude Mythos Preview in April,...
The U.S. government also intervened: on June 12, 2026, Commerce Secretary Howard Lutnick ordered Anthropic to disable Fable 5 and Mythos 5 under an export control order, just three days after their public launch.

Elon Musk and Anthropic logos, representing a dramatic AI industry reversal from public criticism to business partnership. — Search & fact-check with cited sources for What led Elon Musk to publicly reverse his earlier criticism of Anthropic and acknowledge it as tConceptual image: Elon Musk went from calling Anthropic 'misanthropic and evil' to leasing them the world's largest supercomputer in May 2026.
AI Prompt
Create a landscape editorial hero image for this Studio Global article: Search & fact-check with cited sources for What led Elon Musk to publicly reverse his earlier criticism of Anthropic and acknowledge it as t. Article summary: **Important factual note:** The user's question contains several specific claims (Anthropic's "Mythos and Fable models" being named that way, $60 billion ARR, the "180-day" Colossus 1 lease with "mutual 90-day cancellati. Topic tags: general, general web, user generated, news. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts w
openai.com

The Quick Story

In February 2026, Elon Musk called Anthropic's AI models "misanthropic and evil" on X ^F^F^F. Three months later, on May 6, 2026, SpaceX announced a partnership granting Anthropic full access to the Colossus 1 data center in Memphis, Tennessee — the world's largest supercomputer — reportedly valued at roughly $4 billion ^C.

Studio Global AI

Continue your research

This page includes a source-backed answer you can continue inside Studio Global.

Search & fact-check with Studio Global AI

Development	What the Evidence Shows
Mythos & Fable models	Claude Mythos Preview was announced April 7, 2026 ^N. Claude Fable 5 — a guardrailed public version of the Mythos class — launched June 9, 2026 ^T^F^F. Claude Mythos 5 (upgraded, limited-access) followed in July 2026 ^P. Both are "Mythos-class" models that sit above the Opus tier ^H.
$1.2 trillion valuation	Confirmed. On-chain pre-IPO markets and secondary platforms (Forge Global, Hiive) priced Anthropic at a $1.2 trillion implied valuation by early May 2026, surpassing OpenAI ^T^M^C. This was up from a $380 billion primary valuation in February 2026 ^N.
$60 billion ARR	Not confirmed. The highest reported figure is $47 billion annualized run-rate revenue in May 2026 (from Sacra and cited by Forbes/CNBC), up from $9B at end of 2025 and ~$1B at start of 2025 ^F^T^M. A $60 billion figure is not sourced — though a July 2026 SemiAnalysis report projected ARR would surpass $60 billion ^T.
SpaceX Colossus 1 lease	Confirmed as a partnership giving Anthropic the full compute power of Colossus 1 ^C. The lease length (180 days) and mutual 90-day cancellation clause are not detailed in available sources — the Times of India notes Musk "made it clear" he would cancel the deal if Anthropic's AI tools crossed certain lines ^T.
U.S. export controls	Confirmed. On June 12, 2026, just three days after Claude Fable 5's public launch, U.S. Commerce Secretary Howard Lutnick ordered Anthropic to halt access to both Fable 5 and Mythos 5 ^F^T^M. The models were temporarily disabled under an export control order ^N.
Grok 4.5 vs. Claude Opus 4.8	Not found in any source. There is no evidence of models named "Grok 4.5" or "Claude Opus 4.8" in any reporting. xAI was reportedly dissolved and folded into SpaceX AI around May 2026 ^M, which would make direct model-vs-model competition moot at the corporate level.

How Elon Musk Went from Calling Anthropic ‘Evil’ to Powering Its AI — and What It Means for the Industry

Key takeaways

The Quick Story

Continue your research

People also ask

What is the short answer to "How Elon Musk Went from Calling Anthropic ‘Evil’ to Powering Its AI — and What It Means for the Industry"?

What are the key points to validate first?

What should I do next in practice?

Sources

What the Sources Actually Confirm

1. Musk's Earlier Criticism Was Harsh — and Then He Did a Deal

2. No Evidence Musk Called Anthropic the 'Clear AI Leader'

3. Key Developments That Explain the Shift

4. The Deeper Strategic Logic