AIThe Next Webabout 2 hours ago
Open-weight AI nears frontier capability, lags safety
1 min read
GLM-5.2, Z.ai's open-weight model, refused none of the offensive-cyber or dual-use-biology tasks in SaferAI's evaluation, while Claude Opus 4.7 refused so consistently the benchmark could not finish. Open weights cannot be recalled or patched once public, unlike closed models.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
