Anthropic has announced a compute partnership with SpaceX that gives the company access to the full capacity of the Colossus 1 AI data center. The agreement adds more than 300 megawatts of compute power and over 220,000 NVIDIA GPUs, significantly expanding the infrastructure available for Claude models.

The additional capacity is already affecting Anthropic’s products. The company said it is doubling Claude Code’s five-hour rate limits for Pro, Max, Team, and enterprise seat-based plans. It is also removing peak-hour usage reductions for Pro and Max subscribers and substantially increasing API rate limits for Claude Opus models.

According to Anthropic, the SpaceX agreement is part of a broader infrastructure expansion strategy aimed at addressing rising demand for Claude services. The company said the added GPU capacity will directly improve availability for Claude Pro and Claude Max users, who have faced tighter usage restrictions as demand for coding and reasoning workloads increased.

The Colossus 1 facility includes dense deployments of NVIDIA H100, H200, and GB200 accelerators. Anthropic said the compute cluster will support both model training and inference workloads, including Claude Code and API services.

The SpaceX agreement follows several other large-scale infrastructure deals announced by Anthropic this year. These include an agreement with Amazon for up to 5 gigawatts of AI infrastructure, including nearly 1 gigawatt expected online by the end of 2026; a 5 gigawatt partnership with Google and Broadcom beginning in 2027; a strategic infrastructure partnership involving Microsoft and NVIDIA worth up to $30 billion in Azure capacity; and a $50 billion AI infrastructure investment initiative with Fluidstack.

Usage Limits Increase As Demand Surges

The immediate product changes show how tightly compute availability is tied to user experience in large AI systems. Claude Code, which allows developers to use Claude for software engineering workflows, has become one of Anthropic’s most compute-intensive products because coding tasks often require long reasoning chains and repeated iterations.

By raising rate limits and removing peak-hour reductions, Anthropic is effectively signaling that infrastructure constraints had become a bottleneck for paid users. The increase in API capacity also matters for enterprise customers building applications on Claude Opus, Anthropic’s most capable model.

The company’s reliance on multiple hardware platforms, including AWS Trainium chips, Google TPUs, and NVIDIA GPUs, reflects a broader strategy to diversify compute supply instead of depending on a single cloud or chip provider.

AI Infrastructure Expands Beyond The US

Anthropic also said future infrastructure expansion will increasingly happen internationally, particularly for enterprise customers in regulated industries such as healthcare, government, and financial services. Many of these customers require local hosting to meet data residency and compliance rules.

The company said some of its new inference capacity through Amazon will be deployed in Asia and Europe. Anthropic also emphasized that future expansion will prioritize countries with stable legal frameworks and secure supply chains for networking, hardware, and data center infrastructure.

The announcement additionally included continued discussions with SpaceX around orbital AI compute systems. While still experimental, the idea reflects growing concern inside the AI industry that future model development could outgrow the practical limits of terrestrial power, cooling, and land availability.

OpenAI could move into the smartphone market sooner than expected, according to analyst Ming-Chi Kuo, who says the company’s AI-focused handset may enter mass production in the first half of 2027. Earlier reports had pointed to a 2028 timeline, but Kuo now suggests the schedule has accelerated amid rising competition in AI-powered mobile devices and the possibility of an OpenAI public offering.

The device is reportedly being developed around AI-agent functionality rather than traditional smartphone features alone. According to Kuo, MediaTek is expected to become the exclusive supplier of the phone’s system-on-chip. The handset would reportedly use a customized Dimensity 9600 processor built on TSMC’s N2P process technology, which is expected to enter production later this year.

Kuo said the smartphone will place significant emphasis on imaging and real-world visual sensing. The image signal processor is expected to feature an upgraded HDR pipeline designed to improve how the device interprets and processes visual information. The phone is also rumored to include a dual neural processing unit architecture for handling AI workloads directly on the device.

Additional reported specifications include LPDDR6 memory, UFS 5.0 storage, and security technologies such as protected kernel-based virtualization and inline hashing. These features suggest the device is being designed to support more advanced local AI processing while strengthening system-level security.

If development proceeds on schedule, Kuo estimates that cumulative shipments across 2027 and 2028 could reach 30 million units, potentially positioning the device as one of the first large-scale AI-native smartphones from a major AI company.

AI Companies Push Beyond Software Platforms

The reported smartphone project reflects how AI companies are increasingly exploring dedicated hardware to support next-generation AI experiences. Rather than relying entirely on existing mobile ecosystems, firms are looking at ways to integrate AI agents more deeply into operating systems, sensors, and on-device computing infrastructure.

For OpenAI, entering hardware could expand its ecosystem beyond software subscriptions and enterprise services. The company has already been increasing its presence across productivity tools, APIs, and personalized AI assistants, making a dedicated device a logical extension of its platform ambitions.

Microsoft, Google, and xAI have agreed to provide the US government with early access to advanced artificial intelligence models for national security testing. The arrangement will allow federal researchers to evaluate emerging AI systems before public deployment as concerns grow over the cybersecurity and military implications of increasingly capable models.

The agreements were announced by the Center for AI Standards and Innovation, or CAISI, within the US Department of Commerce. The agency said it will use the access to study model capabilities, test for security risks, and assess potential misuse scenarios ranging from cyberattacks to military applications. The move follows a pledge made by the Trump administration in 2025 to establish partnerships with technology companies for AI security evaluations.

Microsoft said it will collaborate with government researchers to test its models for unexpected behaviors and help develop shared datasets and evaluation workflows. The company previously signed a similar arrangement with the United Kingdom’s AI Security Institute. Google declined to comment on the agreement, while xAI did not immediately respond to requests for comment.

The push for earlier government access comes amid rising concern in Washington over the capabilities of frontier AI systems, particularly after Anthropic previewed its Mythos model. Anthropic recently disclosed that Mythos uncovered tens of thousands of software vulnerabilities, raising fears among policymakers and corporate security teams that advanced AI could dramatically accelerate cyberattacks and large-scale hacking operations.

CAISI, formerly known as the US Artificial Intelligence Safety Institute under the Biden administration, has become the government’s primary hub for evaluating advanced AI systems. The agency said it has already completed more than 40 model evaluations, including tests on unreleased systems. In some cases, developers provide versions of models with safety protections partially removed so researchers can more effectively probe for national security risks.

Governments Push For Earlier AI Oversight

The agreements reflect a broader shift toward proactive oversight of advanced AI systems before they are deployed commercially. Policymakers are increasingly concerned that the pace of AI development is outstripping existing regulatory and security frameworks, particularly as models become more capable in coding, reasoning, and autonomous decision-making.

The arrangements also suggest closer coordination between AI developers and national security agencies. As AI becomes strategically important, governments are treating frontier models less like conventional software products and more like critical infrastructure technologies with geopolitical implications.

AI Security Becomes A Strategic Battleground

The agreements build on earlier partnerships established with AI companies during the Biden administration, but they also arrive as competition intensifies among AI developers and cloud providers. The Pentagon recently signed agreements with seven AI companies to deploy advanced systems on classified military networks, signaling growing adoption of AI within defense operations.

Notably, Anthropic was absent from those Pentagon agreements amid reported disagreements over military guardrails for its AI systems. The situation highlights emerging tensions between commercial AI development, government oversight, and defense applications.

As AI capabilities continue to advance, security testing is becoming a core part of the deployment process. Companies are increasingly expected to demonstrate not only model performance, but also resilience against misuse, cyber threats, and unintended behavior before releasing systems at scale.