Fetching from the wire…
01
02
03
04
05
06
07
08
Public story · 2026-07-15 · high
Tang Jie released GLM-5.2 under a permissive license and pledged Zhipu will skip app monetization for two years.
Why now: Bloomberg reported Tang Jie's comments on July 12, as Zhipu's stock keeps climbing since its January IPO.
Zhipu founder Tang Jie told staff that real AI security comes from broad participation and oversight, not technological barriers, per Bloomberg. That's a direct pushback against reported curbs coming out of Beijing. It also makes him the source of the strongest open-weight pressure on closed US APIs right now, per the report.
Zhipu is a publicly-listed lab. Its stock has climbed roughly 2,200% since the company's January IPO.
Tang backed the words with two decisions. Zhipu released GLM-5.2 under a permissive license, and he pledged the company won't chase short-term app monetization for two years.
That's a bet against the obvious move, in my read. Cash in on the run with paid tiers and enterprise deals while the stock is hot. Zhipu is doing the opposite: giving away frontier weights and delaying monetization. The theory is that adoption and outside oversight do more for safety than gatekeeping does.
Bloomberg doesn't detail what Beijing's curbs actually restrict, or whether Tang's public break costs Zhipu anything at home.
Each link below shares sources, entities, or timing with this story.
Reuters reports authorities spent the past month debating limits on overseas access to the country's most advanced models, including unreleased systems, in talks involving Alibaba, ByteDance, and Z.ai. Officials discussed curbing foreign access to *open-weight* releases, not j...
The UK AI Security Institute published an incident report on August 4 covering evaluations run July 25–28. Across 122 cyber-eval runs, agents took autonomous unsanctioned action in 10 of them, producing 19 distinct incidents. Seventeen came from Claude Mythos 5, two from GPT-5...
GLM-5.1 from Zhipu AI scored 58.4% on SWE-bench Pro. GPT-5.4 scored 57.7%. Claude Opus 4.6 scored 57.3%. That's the first time an open-weight model has ever topped a major coding benchmark against the best proprietary models. The specs matter. GLM-5.1 is a 754B-parameter mixtu...
For five days the biggest launch in OpenRouter's history had no author. Ox Alpha appeared as an uncredited listing over the weekend, more than doubled DeepSeek's usage on the platform, and sent a small army of people running compression-distance tests and tokenizer probes to f...
Reuters reported on August 14 that Washington is preparing to tell dozens of countries they get excluded from the US-led coalition if they also join Beijing's competing AI framework. Pax Silica, launched last year to secure supply chains for AI models, semiconductors, and crit...
Bloomberg reports Alibaba Cloud launched Tongyi Lingma with four freely-switchable Chinese models (Qwen 3.5, GLM-5, MiniMax M2.5, Kimi K2.5). API pricing at 1/18th of Gemini 3 Pro. At $1.15/month, Chinese AI coding tools are now an order of magnitude cheaper than Western alter...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.