AI News Summary — Week 20.1, 2026

This week: Anthropic goes big on compute with SpaceX, OpenAI brings its cyber defense model to Europe, and the AI infrastructure race reaches new heights. Let's dive in.

Anthropic Secures SpaceX Compute Deal — 220,000 New GPUs

Anthropic just landed a massive compute partnership with SpaceX that will substantially boost their AI capacity. Under the agreement, Anthropic will get access to all compute capacity at SpaceX's Colossus 1 data center — that's more than 300 megawatts of power and over 220,000 NVIDIA GPUs, available within the month.

This is just the latest in a series of compute deals for Anthropic:

Amazon: Up to 5 gigawatts agreement, including nearly 1 GW of new capacity by end of 2026
Google & Broadcom: 5 GW coming online starting in 2027
Microsoft & NVIDIA: $30 billion of Azure capacity
Fluidstack: $50 billion investment in American AI infrastructure

The SpaceX deal means Anthropic can now double Claude Code's five-hour rate limits for Pro, Max, Team, and Enterprise plans, and remove peak hours limit reductions for Pro and Max accounts. API rate limits for Claude Opus models have also been raised considerably.

Orbital Compute?

As part of the agreement, Anthropic has also expressed interest in partnering with SpaceX to develop multiple gigawatts of orbital AI compute capacity — yes, that's AI compute in space.

OpenAI Expands GPT-5.5-Cyber to Europe

OpenAI announced an EU Cyber Action Plan this week, expanding access to GPT-5.5 and GPT-5.5-Cyber for trusted European cybersecurity defenders. The company is rolling out GPT-5.5-Cyber in limited preview to defenders responsible for securing critical infrastructure.

The Trusted Access for Cyber (TAC) framework now has three levels:

GPT-5.5 (default): Standard safeguards for general-purpose use
GPT-5.5 with TAC: More precise safeguards for verified defensive work — includes secure code review, vulnerability triage, malware analysis, detection engineering, and patch validation
GPT-5.5-Cyber: Most permissive behavior for specialized authorized workflows — includes authorized red teaming, penetration testing, and controlled validation

Starting June 1, 2026, individual TAC users accessing the most cyber-capable models will be required to enable Advanced Account Security with phishing-resistant authentication.

What's Hot This Week

Compute is king: Anthropic's SpaceX deal gives them 220,000 GPUs in a month — the infrastructure race is accelerating
Orbital AI compute: Anthropic interested in SpaceX partnership for space-based AI — sci-fi is becoming reality
Cyber defense goes global: OpenAI's EU expansion signals that AI cybersecurity is becoming a geopolitical priority
Rate limits increasing: Claude Code users get doubled limits and no more peak hour reductions — enterprise adoption is scaling

That's a wrap for this week. Stay tuned for more AI developments!

Sources: Anthropic Blog, OpenAI Blog, TechCrunch, The Verge, CNBC