DeepSeek launches 1.6 trillion parameter V4 on Huawei chips as U.S. escalates AI theft accusations — U.S. gov't alleges IP theft by DeepSeek and other Chinese AI firms

DeepSeek previewed V4, its largest model yet, with a 1 million-token context window and low-cost pricing.

The company said V4 is optimized for Huawei’s Ascend chips, signaling a shift away from Nvidia hardware.

On the same day, Reuters reported U.S. guidance to embassies to raise concerns about alleged AI model extraction and distillation by DeepSeek and other Chinese firms.

The allegations add to broader U.S.-China tensions over AI intellectual property, training data, and export controls.