The US government ordered federal agencies to remove DeepSeek technology and extended restrictions to third-party products built using its AI models.
On Oct. 20, the Department of Homeland Security (DHS) published the directive, requiring federal civilian agencies to eliminate the Chinese company’s products from their information technology systems and prohibiting future purchases.
The restrictions also cover software and services developed by other companies using DeepSeek implementations, potentially affecting American technology contractors that have incorporated its models into their products.
Homeland Security Secretary Markwayne Mullin originally signed the order on July 24, under the Federal Acquisition Supply Chain Security Act of 2018. Agencies were required to begin complying immediately, with a 60-day window to request exceptions.
DHS cited national security concerns about DeepSeek’s Chinese operations, including the possibility that government employees could enter sensitive information that could be transmitted to servers in China.
The department also warned that the company’s technology could contain security vulnerabilities and that information collected through its services could provide intelligence on American government activities.
The restrictions expand Washington’s efforts to limit Chinese technology in government infrastructure, following earlier actions against DeepSeek by the Commerce Department, the US Navy, and several state governments.
China condemned the decision, accusing Washington of politicizing technology and undermining commercial cooperation between the world’s two largest economies.
The order creates new compliance obligations for federal technology suppliers, which must assess whether products incorporating DeepSeek implementations can continue serving government customers. Affected contractors may need to replace those components or seek exemptions, potentially benefiting competing AI providers.
DeepSeek opens up AI technology
Against that backdrop, DeepSeek announced on Oct. 20 that it had open-sourced DeepInfer, the inference framework powering its large-scale artificial intelligence services, giving developers access to technology previously used internally to operate its models.
The release includes optimized computing kernels, distributed inference capabilities, speculative decoding, and processing tools designed to improve throughput and reduce the computing resources required to generate AI responses.
DeepSeek also released deepseek-recipe, a collection of software components that handle token encoding, prompt formatting, and API request processing. The tools let independent developers reproduce key elements of the company’s API behavior when running its models.
“We are releasing DeepInfer, deepseek-recipe, and other vital components for serving DeepSeek models,” company engineer Cui Tianyi said, adding that the goal was to enable more businesses and organizations worldwide to deploy and profit from the technology.
The move expands DeepSeek’s open-source strategy beyond publishing model weights, giving developers greater control over the infrastructure they use to run AI systems.
That could strengthen the company’s appeal among businesses seeking to reduce reliance on proprietary AI services from providers such as OpenAI, Anthropic, and Google, whose flagship production-serving infrastructure remains largely closed.
DeepInfer currently supports DeepSeek-V4.1-Flash, a 552-billion-parameter model introduced in September that activates only a fraction of its parameters for individual tasks.
The company said the model’s architecture requires substantially less memory and storage for inference caching, allowing providers to process more requests using fewer computing resources.
DeepInfer is distributed under the Apache 2.0 license, permitting commercial modification and redistribution. It supports Nvidia SM100 graphics processors and Huawei Ascend 950 chips, giving developers options across American and Chinese computing hardware.
The compatibility with Huawei’s processors could also support Chinese efforts to develop domestic AI infrastructure amid US restrictions on advanced semiconductor exports.
The initial release provides offline inference capabilities, although operators must still supply components such as cache management and load balancing to build a complete production service.
Ultimately, the release allows cloud providers and businesses outside the US federal market to run DeepSeek models on their own infrastructure, tailor the technology to their needs, and reduce AI operating costs.
However, hosting DeepSeek technology on private infrastructure would not automatically exempt a product from the federal order, which expressly covers third-party implementations.
Credit: Source link