DHS bans DeepSeek AI across US federal civilian agencies

By Marcus Bennett

The US Department of Homeland Security has banned federal civilian agencies from using AI technology from Chinese lab DeepSeek, according to a report published by The Decoder on June 2026-dated coverage of the directive.

The decision is the strongest concrete signal to date that Washington is moving from scrutiny to outright prohibition of DeepSeek's models inside the federal government. The ban covers civilian agencies, which places it in the same category of restriction the US has previously applied to foreign technology deemed a security concern.

What does the ban actually cover?

The Department of Homeland Security, the federal department responsible for domestic security policy, issued the prohibition against AI technology from DeepSeek, the Chinese AI laboratory whose models have drawn international attention since their public release.

The directive applies to federal civilian agencies. That scope matters. Civilian agencies handle records, communications and operational data across the breadth of the federal workforce, and a prohibition at that level effectively removes DeepSeek's technology from the day-to-day tooling of non-defense government bodies.

The Decoder, which first reported the ban, stated the core fact plainly: the Department of Homeland Security "has banned federal civilian agencies from using AI technology from Chinese lab Deepseek."

Why does DeepSeek draw this level of restriction?

DeepSeek is a Chinese AI lab. That single designation carries substantial weight in current US policy, where regulators and lawmakers have treated Chinese-developed AI systems as potential vectors for data exposure and foreign influence.

For federal agencies, the stakes are concrete. Agency staff who use an AI model in the course of their work can expose sensitive questions, documents and internal processes to the system's operator. A ban removes that exposure path entirely, rather than managing it through usage policies.

This is the context in which the Homeland Security decision lands. The US government has spent the past several years restricting Chinese technology from federal networks, and AI models from Chinese labs have now entered that same policy perimeter.

What happens next?

The immediate consequence is removal. Federal civilian agencies that had access to or were evaluating DeepSeek technology now face a directive that forecloses its use.

The longer consequence is precedent. A department-level ban from Homeland Security establishes a template that other federal bodies, and potentially state governments, can follow against AI systems from labs the US considers a security risk. Watch for whether the restriction widens beyond civilian agencies or extends to additional Chinese AI providers in the coming months.

---

This report is based on coverage published by The Decoder. The source report, titled "US bans Deepseek technology across federal civilian agencies," states that the US Department of Homeland Security has banned federal civilian agencies from using AI technology from Chinese lab DeepSeek. Additional details on the scope, effective date and implementation of the ban were not disclosed in the source material and will be reported as they emerge.