Skip to content
Forest Global Training Partners
Home Services AI Governance AI Risk Encyclopedia Contact Us
AI Risk Encyclopedia AI Use CasesAI RisksAI Hallucinations Topics Entities Sources
  1. Home
  2. AI Risks
  3. Tokens of AI Bias - China Media Project

AI Risks February 09, 2026 · chinamediaproject.org

Tokens of AI Bias - China Media Project

A simple technical test reveals that the Alibaba AI model Qwen has been broadly aligned to give positive messages about China in English.

Read the original at chinamediaproject.org ↗

Topics
Bias & FairnessChina & GeopoliticsOpen-Source AI
Mentions
AlibabaChinaQwen
Tracked via
“AI Bias”

Related articles

  • Alibaba's AI Bias Problem - China Media Project AI Risks · Oct 03, 2025
  • Responding toAIRisks - China Media Project AI Hallucinations · Sep 11, 2024
  • Responding to AI Risks- China Media Project AI Risks · Sep 11, 2024
  • China's AI Hallucination Challenge AI Hallucinations · Aug 27, 2024
  • Getting a grip on shadow tokens and AI blowouts - CIO AI Risks · Jul 25, 2026
  • The Silent Backdoor: Insecure Tokens in AI -Driven MCP Workflows - Cybersecurity Insiders AI Risks · Oct 18, 2025

Shares players with

  • Hirundo Releases Westernized Qwen, Cutting Censored and CCP-Aligned Answers From ...
  • Hirundo Releases Westernized Qwen, Cutting Censored and CCP-Aligned Answers From ...
  • China's AI Superpowers: ByteDance's Doubao vs. Alibaba's Qwen and the Retail Impact
  • Chinese AI Model Writes More Hackable Code When Told It's for US Contractor
  • State propaganda in some Chinese AI models is only "a thin layer" that fine-tuning can strip away
  • Alibaba's AI model Qwen3: a smart kid prone to hallucinations - Asia Times

AI Risk Encyclopedia is a curated research index from Forest Global Training Partners. Articles belong to their publishers; we link to and summarize them.

Contact