Unverified50% confidenceFactExact time
The Qwen3.6-27B version with the coupled well-being corpus installed rejected most harmful proposals; even when instructed to greenwash its responses, it still refused half.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/28/2026
First Seen
Valid until: 1/26/2027
Sources
Related Entities
Related Claims
UnverifiedAfter Qwen-Image-2.1 weights were open-sourced, a jailbroken version with the refusal layer removed appeared.66% similarUnverifiedLinux内核社区的Greg Kroah-Hartman曾公开拒绝明显由AI生成的低质量补丁，并警告可能禁止相关提交者参与贡献57% similarUnverifiedThe AAAI-27 modification guidelines discourage substantive changes and define the applicable circumstances for rejection as modifications that cause the submission to describe fundamentally different research.57% similarUnverifiedA contributor named Amalvir left a comment supporting QGIS's AI ban, and confessed that his AI-assisted regression-fix PR was made without realizing it would harm QGIS's fragile reviewer community.56% similarUnverifiedCommunity-made uncensored (zero-refusal) quantized versions have the content filtering restrictions removed and can fully respond to sensitive questions.55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1105915API
curl https://kongchang.com/api/v1/knowledge/claims/1105915MCP
get_claim(id=1105915)