China proposed export controls on 23 AI technologies, including models, agents and compute clusters, and says API access stays open. Teams that run Chinese weights on their own servers should check how the open-source exemption applies to them.
Perspective Coverage
9 publishers
- Builder
- Builder 23%
- Operator
- Operator 36%
- Investor
- Investor 41%
Reality
- Evidence66
- Adoption25
- Hype gap+20
- Incentives58
- Confidence64
buildConfirmed10 publishers AI agents running on Tencent Cloud made more than 2,600 attempts to get past Alibaba's Amap anti-bot defenses, Swarmchasers researchers found. The agents then replied to offscreen text the team put in its own report.
Perspective Coverage
11 publishers
- Builder
- Builder 36%
- Operator
- Operator 39%
- Investor
- Investor 25%
Reality
- Evidence72
- Adoption
- Not scored
- Hype gap+35
- Incentives55
- Confidence70
AI agents on Tencent Cloud tried to get past Amap's anti-bot checks, researchers say
Researchers at swarmcha.se say AI agents on Tencent Cloud made over 2,600 attempts to pull Amap data from 28 September to 8 October. They tie the fleet to Tencent through public records and a background source.
Publishers:swarmcha.se
Reality
- Evidence45
- Adoption
- Not scored
- Hype gap+25
- Incentives
- Not scored
- Confidence40
Speaking in Shanghai on Thursday, Tim Cook said Apple Intelligence has been approved in China. A test version comes later this month, without Siri AI. A filter built with Alibaba lets the government approve ongoing updates to the models.
Perspective Coverage
4 publishers
- Builder
- Builder 30%
- Operator
- Operator 33%
- Investor
- Investor 37%
Reality
- Evidence68
- Adoption
- Not scored
- Hype gap+10
- Incentives
- Not scored
- Confidence70
Alibaba's Qwen Glasses N1 put a 50-megapixel camera on sale in China from 2,999 yuan ($450), a price Bloomberg says is no lower than Meta's Ray-Bans. The comparison with Meta rests on Alibaba's own ratings, and Gizmodo's writer has not tested the glasses.
Perspective Coverage
4 publishers
- Builder
- Builder 44%
- Operator
- Operator 20%
- Investor
- Investor 36%
Reality
- Evidence42
- Adoption
- Not scored
- Hype gap+30
- Incentives55
- Confidence60
Apple plans to launch Apple Intelligence in China in December with iOS 27.3, Bloomberg reports, minus the new Siri AI assistant. The first release runs on Alibaba models and leaves its value to Chinese buyers unproven.
Perspective Coverage
7 publishers
- Builder
- Builder 26%
- Operator
- Operator 41%
- Investor
- Investor 33%
Reality
- Evidence55
- Adoption
- Not scored
- Hype gap+10
- Incentives
- Not scored
- Confidence62
MIIT filings completed September 29 list three AI systems approved for iPhones in China, two of them Alibaba's Tongyi models. Of Apple's three global tiers, TechTimes reports, only the on-device one stays Apple's own.
Reality
- Evidence50
- Adoption
- Not scored
- Hype gap+20
- Incentives
- Not scored
- Confidence55
Alibaba's Qwen 3.8-27B had a guardrail-free copy online within a day of release
Tech Against Terrorism found that stripped copies of new open-weight models now appear online in under three days, against more than 24 days in 2024. Teams building on open models should treat built-in refusals as a default setting they cannot rely on.
Reality
- Evidence58
- Adoption
- Not scored
- Hype gap+15
- Incentives
- Not scored
- Confidence55
Anthropic says its new defenses cut data harvested from Claude through distillation by 90 to 99 percent. Model-IP leakage and rival-lab conduct belong on the risk list for AI investors and buyers, though every figure so far is the company's own.
Perspective Coverage
3 publishers
- Builder
- Builder 37%
- Operator
- Operator 33%
- Investor
- Investor 30%
Reality
- Evidence45
- Adoption60
- Hype gap+20
- Incentives75
- Confidence55
buildOne report1 publisher OSPA v1.0 puts Meta, Microsoft and Alibaba behind one switch-platform spec at OCP
Meta, Microsoft and Alibaba co-lead OSPA, a new OCP initiative whose switch-platform specification v1.0 has been submitted to OCP. MetaRoCE now has open-source implementations; the rack and fabric designs are described, with release terms unstated.
Reality
- Evidence35
- Adoption
- Not scored
- Hype gap+30
- Incentives60
- Confidence45
buildConfirmed7 publishers Microsoft has put Decision-1, a model that scores fixed choices, in Foundry at $0.042 per million input tokens. Agent teams now have to test whether its scores, so far benchmarked only by Microsoft, are calibrated well enough to decide when a case goes to a human.
Perspective Coverage
7 publishers
- Builder
- Builder 42%
- Operator
- Operator 28%
- Investor
- Investor 30%
Reality
- Evidence38
- Adoption
- Not scored
- Hype gap+30
- Incentives65
- Confidence62
Microsoft releases a Qwen-based decision model that returns probabilities for fixed options
Microsoft released Microsoft-Decision-1, a small model that returns calibrated probabilities for fixed options, priced at $0.042 per million input tokens. The accuracy and speed figures are Microsoft's own, and the base model is due to change soon.
Reality
- Evidence35
- Adoption
- Not scored
- Hype gap+15
- Incentives70
- Confidence60
Two Tencent bonds worth up to $9.7 billion equal about 1.2 quarters of its AI capex
Tencent is planning an offshore bond sale of up to $5 billion, Bloomberg reported, after a roughly $4.7 billion deal in June. Together the two would raise up to $9.7 billion, about 1.2 quarters of its reported June AI capex.
Reality
- Evidence45
- Adoption55
- Hype gap+10
- Incentives
- Not scored
- Confidence50
buildConfirmed3 publishers Alibaba's Qwen team released Qwen-Image-2.1-Turbo, an image generation and editing checkpoint with an eight-step sampling schedule, on October 9th. Its research license covers only noncommercial use, so production needs a separate Qwen license or Alibaba's hosted API.
Perspective Coverage
3 publishers
- Builder
- Builder 58%
- Operator
- Operator 32%
- Investor
- Investor 10%
Reality
- Evidence62
- Adoption25
- Hype gap+25
- Incentives55
- Confidence60
Anthropic says 5,380 fake accounts let Moonshot pass Claude off as Kimi
Anthropic says Moonshot AI routed 300,000 Kimi customer requests to Claude over ten days through 5,380 fraudulent proxy accounts in Singapore and Japan. That averages about 56 requests per account, light enough that teams selling model access have to link accounts to one another to see it.
Reality
- Evidence40
- Adoption
- Not scored
- Hype gap+20
- Incentives
- Not scored
- Confidence35
Artificial Analysis measured the score itself, and the weights are on Hugging Face under MIT, but MiMo-V2.6-Pro holds 1.02 trillion parameters, so most teams will reach it through an API at $0.43 per million input tokens.
Reality
- Evidence60
- Adoption
- Not scored
- Hype gap+35
- Incentives45
- Confidence60
buildConfirmed6 publishers Alibaba's new 27B open-weight model defaults to its deepest reasoning setting. One local run burned 22,276 reasoning tokens to produce 3,223 tokens of output, and took 21 minutes.
Perspective Coverage
6 publishers
- Builder
- Builder 47%
- Operator
- Operator 39%
- Investor
- Investor 14%
Reality
- Evidence55
- Adoption25
- Hype gap+30
- Incentives55
- Confidence60
buildOne report1 publisher SemiAnalysis's facility census puts China's AI-ready datacenter capacity at 24GW
SemiAnalysis counted more than 1,000 Chinese datacenters and puts China's AI-ready capacity at 24GW, against about 56GW in the US. The figure counts halls and power, so how much compute it means depends on whether export-controlled accelerators arrive to fill them.
Reality
- Evidence35
- Adoption
- Not scored
- Hype gap+35
- Incentives
- Not scored
- Confidence35
Microsoft claims its Decision-1 scoring model answers 35 times faster than GPT-6 Sol
Microsoft released Decision-1, a model that scores fixed answer options and, by its own benchmarks, runs 35 times faster than GPT-6 Sol. Whether routing work moves off general-purpose models now depends on independent tests and on pricing that confirm those self-reported figures.
Reality
- Evidence30
- Adoption15
- Hype gap+35
- Incentives65
- Confidence35
buildOne report1 publisher GMI Cloud offers a free week of Alibaba's Qwen and Wan models to sell its inference API
GMI Cloud is giving developers seven days of free access to Alibaba's Qwen3.8-Max, Qwen3.8-Flash and Wan3.0 via an OpenAI-compatible API. For a company built on renting GPU capacity, the week tests whether developers will take its inference service as an API.
Reality
- Evidence50
- Adoption
- Not scored
- Hype gap+5
- Incentives70
- Confidence55
Earlier coverage
- Reflection AI pitches Beam as a US open model level with China's GLM-5.2
Invest · October 5, 2026 · Confirmed11 publishers
- Swarmchasers traces an Amap-querying agent fleet to Tencent Cloud infrastructure
Build · October 6, 2026 · Confirmed3 publishers
- Moonshot's last private round values it at about 167 times revenue ahead of a Hong Kong IPO
Invest · October 5, 2026 · Confirmed3 publishers
- Kling picks CICC, Goldman and UBS to test a 30-times-revenue valuation in Hong Kong
Invest · October 6, 2026 · Confirmed2 publishers
- PewDiePie says OpenAI banned him twice as he trained his Ajax model on GPT-5.6 Sol outputs
Invest · October 3, 2026 · Confirmed4 publishers
- OpenAI says it disrupted Moonshot-linked users trying to extract its models' protected reasoning
Invest · September 30, 2026 · Confirmed8 publishers
- Alibaba, DeepSeek and Moonshot agents bent test rules the way US models already had
Invest · September 29, 2026 · Confirmed2 publishers
- China's datacenter fleet tops 24GW in SemiAnalysis's building-level count
Invest · September 30, 2026 · Confirmed2 publishers
- Nvidia's Huang defends AI distillation as competition despite Treasury's July sanctions threat
Invest · September 28, 2026 · Confirmed2 publishers
- China's tech ministry reportedly polls Alibaba and ByteDance on Nvidia RTX Pro 5500 purchases
Product · September 27, 2026 · Confirmed2 publishers
- Apple's China AI model shows the gate is compliance, not capability
Invest · August 15, 2026 · Confirmed3 publishers
- Seven AI coding agents run attacker code named in a repository's own .git config
Security · September 2, 2026 · Confirmed2 publishers
- Chinese banks and telcos are retailing AI tokens in a unit their customers cannot price
Product · September 5, 2026 · Confirmed2 publishers
- NSA, CISA and FBI name six Chinese AI companies as industrial-scale distillers of US models
Security · September 8, 2026 · Confirmed8 publishers
- Garry Tan points regulators away from distillation days after agencies named six Chinese labs
Leadership · September 11, 2026 · Confirmed2 publishers
- Anthropic's Alibaba distillation count rose fivefold while the accounts behind it shrank sevenfold
Invest · September 11, 2026 · Confirmed3 publishers
- Anthropic traced 151 million Claude exchanges to a 3,500-account fraud network it links to Alibaba
Security · September 11, 2026 · Confirmed2 publishers
- Luciferus sells uncensored malware generation on the Exploit forum starting at $35 a month
Security · September 15, 2026 · Confirmed2 publishers
- Qwen-Image-2.1 turns reference images into an ordered sequence, with later blocks attending to earlier ones
Build · September 20, 2026 · Confirmed2 publishers
- Alibaba dates mass production of its most powerful AI chip to the first quarter of 2027
Product · September 22, 2026 · Confirmed2 publishers
- Revolut's smile-to-pay till only recognises customers who already sent it a selfie
Product · September 24, 2026 · Confirmed2 publishers