Welcome back to Daily Zaps, your regularly-scheduled dose of AI news ⚡

Here’s what we got for ya today:

Let’s get right into it!

Sign Up | Advertise | Tools

Anthropic has agreed to use the entire compute capacity of SpaceX's Colossus 1 data center in Memphis — 300+ megawatts, more than 220,000 NVIDIA GPUs (H100, H200, and next-gen GB200s) — coming online within the month. SpaceX acquired xAI earlier this year, making this a deal between the two biggest non-OpenAI AI players in the world. It's also the largest single compute commitment Anthropic has announced with a fellow AI-adjacent company.

The capacity shows up in your account today. Claude Code's five-hour rate limits are doubling for Pro, Max, Team, and Enterprise plans. Peak-hours throttling is being removed for Claude Code on Pro and Max. API rate limits for Claude Opus models are rising considerably. This joins Anthropic's broader compute buildout: 5GW with Amazon, 5GW with Google and Broadcom, and $30B in Azure capacity with Microsoft and Nvidia. The company also disclosed interest in developing multiple gigawatts of orbital AI compute with SpaceX — AI infrastructure, in space.

📖 Read the announcement at Anthropic →

The lawsuit alleged Apple marketed Siri as having advanced AI capabilities it announced at WWDC 2024 — and never shipped. The promised features: a deeply personalized Siri that could take actions inside apps, understand personal context, and integrate with third-party services. None of it arrived on time. A class action followed.

Apple has agreed to a $250 million settlement covering roughly 36 million eligible devices: iPhone 15 Pro, 15 Pro Max, and iPhone 16 models purchased in the U.S. between June 10, 2024 and March 29, 2025. Each eligible device could yield between $25 and $95 depending on how many people file claims — the fewer who apply, the more each claimant receives. The settlement still needs court approval at a June 17 hearing. Apple does not admit wrongdoing.

📖 Full breakdown

Parag Agrawal — fired as Twitter's CEO the day Elon Musk completed his acquisition — just closed a $100 million Series B for Parallel Web Systems at a $2 billion valuation, led by Sequoia. That's a 2.7x valuation jump from his $740 million Series A just five months ago. Total raised: $230 million.

Parallel is building programmatic web infrastructure specifically for AI agents: APIs for searching the internet, extracting data from websites, completing tasks online, and monitoring the web for changes. The problem it solves: today's AI agents are bottlenecked by a web built for human browsers — slow, unstructured, and locked behind interfaces agents can't use. Customers already include Clay, Harvey, Notion, and Opendoor, plus undisclosed banks and hedge funds, with over 100,000 developers building on its APIs today.

📖 Read the full story

The Defense Department has reached agreements with eight companies — Amazon Web Services, Google, Microsoft, Nvidia, OpenAI, SpaceX, Reflection, and Oracle — to deploy AI tools on its IL6 and IL7 classified network environments. These are the highest-security tiers in U.S. military infrastructure, housing some of the most sensitive intelligence and operational data in the country.

The stated goals: improve warfighter decision-making, data synthesis, and situational awareness. The Pentagon says it's building "an architecture that prevents AI vendor lock-in" — keeping options open as the AI market shifts. The agreements don't disclose dollar amounts, specific use cases, or deployment timelines, but the scope is significant: these are the same classified networks used for top-secret operations. The fact that commercial AI companies now have direct access to them marks a meaningful shift in how the U.S. military is approaching AI adoption.

📖 Official Pentagon announcement →