# Chinese AI firms offer up to 1.09 million yuan for deployment engineers as export-barred Nvidia B300 chips appear on China Telecom-linked systems

Postings for forward-deployed engineers in China rose 1,523% year on year, while a scan of unsecured GPU monitoring endpoints found 32 Blackwell Ultra chips that United States rules bar from sale to China.

- Published: 2026-10-26T06:26:37.653Z
- Canonical: https://polylog.news/ai/2026-10-26/chinese-ai-firms-offer-up-to-1-09-million-yuan-for-deploymen
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Polylog editors](https://polylog.news), [South China Morning Post](https://www.scmp.com/tech/tech-trends/article/3371925/chinas-tech-giants-are-paying-1-million-yuan-new-kind-ai-engineer), [FourWeekMBA](https://fourweekmba.com/ai-bytedance-offers-up-to-65000-yuan-a-month-for-ai-deployers/), [Crypto Briefing](https://cryptobriefing.com/lava-nvidia-b300-chips-china-telecom/), [DigiTimes](https://www.digitimes.com/news/a20261023VL220/nvidia-chips-security-blackwell-china-telecom.html), [Yahoo Finance](https://finance.yahoo.com/technology/ai/articles/u-export-restrictions-drive-nvidia-124325286.html), [The Next Web](https://thenextweb.com/news/nvidia-b300-server-1-million-china-export-controls)

ByteDance, Tencent and Alibaba are offering close to 1 million yuan (about $150,000) a year for forward-deployed engineers, who build and integrate AI systems inside client companies, [the South China Morning Post reported](https://www.scmp.com/tech/tech-trends/article/3371925/chinas-tech-giants-are-paying-1-million-yuan-new-kind-ai-engineer). Moonshot's Kimi advertised 1.09 million yuan ($163,000), according to [AI Post](https://t.me/aipost/8614). One ByteDance listing offers 975,000 yuan a year, structured as 65,000 yuan a month paid 15 times, and the company's deployment role for its Doubao model pays 35,000 to 70,000 yuan monthly, [FourWeekMBA reported](https://fourweekmba.com/ai-bytedance-offers-up-to-65000-yuan-a-month-for-ai-deployers/). Postings for the role rose 1,523% from a year earlier.

The premium follows from how Chinese labs earn revenue. Their models are priced low or distributed as downloadable weights, so money comes from getting those models into client workflows. That integration work is labor, and labor falls outside American export rules.

Compute does not. Lava, an Israeli security firm, identified 32 Nvidia Blackwell Ultra B300 GPUs reporting through monitoring endpoints linked to state-owned China Telecom, [Crypto Briefing reported](https://cryptobriefing.com/lava-nvidia-b300-chips-china-telecom/). The B300 is barred from sale to China. Lava found the chips through publicly reachable instances of Nvidia's Data Center GPU Manager exporter that required no authentication, part of roughly 2,100 exposed hosts worldwide reporting more than 12,000 GPUs across nearly 300 organizations. Lava could not establish how the chips arrived or where in the network they sit, [DigiTimes noted](https://www.digitimes.com/news/a20261023VL220/nvidia-chips-security-blackwell-china-telecom.html).

The controls still bind. Nvidia's share of China's data-center AI chip market fell to 34% in the first half of 2026, according to [an International Data Corporation report cited by Yahoo Finance](https://finance.yahoo.com/technology/ai/articles/u-export-restrictions-drive-nvidia-124325286.html), with domestic suppliers taking the difference. Earlier this year B300 servers sold in China for roughly $1 million each, close to double the United States price, [according to The Next Web](https://thenextweb.com/news/nvidia-b300-server-1-million-china-export-controls). Nvidia says the B300 is restricted from sale in China and that its partners must comply strictly with export rules.

## What this means

Export controls raise the price of Chinese compute without closing off the supply. Smuggling roughly doubles server cost and pushes buyers toward domestic accelerators, so Chinese labs compete where the rules do not reach: on model price and on paid deployment labor. Nvidia is exposed through lost China market share and compliance scrutiny each time grey-market hardware surfaces. Domestic Chinese accelerator vendors gain by default. Server assemblers and Southeast Asian transit intermediaries face the greatest risk of enforcement attention. The common claim that the controls have failed overstates the case. A 34% market share and a doubled local price are themselves evidence that the controls are working, even as some chips get diverted around them. The precedent is the 2025 and 2026 diversion prosecutions, including Taiwanese charges against nine people over AI server exports, which produced tighter reseller scrutiny rather than new chip categories. The likelier next step over the coming months is pressure for hardware location verification and enforcement aimed at cloud providers and resellers rather than at smugglers alone, conditional on Washington holding its current trade posture. The read is wrong if the administration broadly licenses cut-down Blackwell parts for China, which would remove the price gap that funds diversion. For investors, the pressure point is rising engineering wages squeezing margins at Chinese AI services companies.

## What to watch

- Whether the United States Commerce Department opens an inquiry following Lava's finding, which would show enforcement shifting from smugglers to end users and their suppliers.
- Nvidia's next quarterly disclosure of China revenue and export-license status, the cleanest measure of what the controls cost the company.
- Whether Chinese deployment-engineer pay keeps climbing, since sustained increases would mean model margins are being spent on human integration work.