← The Polylog AI Intelligence Brief
Morning Edition · Monday, October 26, 2026Published at 2:26 AM EDT · New York
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.
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. Moonshot's Kimi advertised 1.09 million yuan ($163,000), according to AI Post. 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. 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. 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.
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, 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. Nvidia says the B300 is restricted from sale in China and that its partners must comply strictly with export rules.
Veracity: Plausible
72/100Part of a tracked trend
AI Sovereignty and Export Controls on Frontier Models
Over the next 3-6 months, governments increasingly treat frontier AI models as strategic national assets — extending export controls to model access itself and backing domestic 'champion' labs as sovereignty plays.
Townsquare
Start a discussion in Townsquare.
More from this edition
- Recursive releases Flint, a free coding model with 2.5 billion active parameters that it says its own AI research system mostly built
- Anthropic says a Claude research model designed 318,484 protein candidates in 64 hours, and 17 passed laboratory binding tests
- Anthropic prices Claude Opus 5.5 at 20% below Opus 5 per token and claims 66.4% on Terminal-Bench 4.0
- Berkeley Lab cut beamline image analysis from months to about 15 minutes using Meta's open-weight vision models
- OpenAI's Decisions API bills input tokens only and returns a typed answer in about 150 milliseconds
- A mathematicians' association has 1,008 members pledged not to work with commercial AI companies
- Apollo's chief economist says AI has produced no financial payoff for United States companies outside technology
Comments
0No comments yet.