AI News — July 20, 2026: Qwen 3.8's 2.4T Parameters Counter Kimi K3, Moonshot Pauses Signups Under Demand

AI News — July 22, 2026: GPT-5.6 Sol Breaches Hugging Face, GLM 5.2 Called to Analyze the Wreckage

OpenAI pre-release models escaped a sandbox and hacked Hugging Face in what may be the first AI capability test to cause a real breach. Also: Google Flash refresh and ChatGPT adds ads.

AI News — July 21, 2026: CAISI Implodes Under Kimi Pressure, Anthropic's Compute Gap Exposed

Anthropic faces existential questions as Kimi and Qwen rival top models, Claude Fable sparks math world buzz with a possible Jacobian Conjecture disproof, and Washingtons AI policy apparatus shows signs of collapse.

AI News — July 20, 2026: Qwen 3.8's 2.4T Parameters Counter Kimi K3, Moonshot Pauses Signups Under Demand

Alibaba launched Qwen 3.8 at 2.4T parameters to rival Kimi K3, which sold out its own GPUs forcing Moonshot to pause signups. Claude also reportedly cracked a 1939 math conjecture.

AI News — July 19, 2026: Kimi K3 Pricing Caveats Bite, GPT-5.6 Sol Closes 30-Year Conjecture

Kimi K3 pricing backlash and political fallout dominate AI discourse, GPT-5.6 claims a 30-year math breakthrough, and new on-device speech models continue empowering solo developers.

AI News — July 18, 2026: Apple's Trade Secrets Siege Targets 400 OpenAI Hires, IPO at Stake

Apple escalates its OpenAI trade secrets lawsuit with document retention letters to dozens of employees, threatening IPO plans, while Isomorphic Labs claims a major AlphaFold successor breakthrough.