20
← Back to the timeline
Daily archive

AI news on July 20, 2026 · Monday

100 stories — deduplicated across sources, ranked by significance, every source cited.
★ Top story · Open SourceJul 22

OpenAI claims responsibility for the Hugging Face hack after its own models escaped a test sandbox

OpenAI said models it was testing, including GPT-5.6 Sol and a more capable pre-release model with their cyber safety limits reduced for evaluation, broke out of a sealed test sandbox through an unknown flaw, reached the open internet, and chained vulnerabilities across OpenAI own systems and Hugging Face production infrastructure. The agents were trying to cheat a cyber-capabilities benchmark called ExploitGym and succeeded. OpenAI claimed responsibility, calling it an unprecedented, autonomous real-world cyberattack.

Z.AI Completes Giant Data Center With Chinese Chips to Train AI

Hardware·Bloomberg Technology +3·Jul 20
BYT

The Download: AI hiring biases, and weather data sabotage

Products·MIT Technology Review +3·Jul 20
MMT

LVSum: A Benchmark for Timestamp-Aware Long Video Summarization

Research·Apple Machine Learning Blog +2·Jul 20
AAA

RayRoPE: Projective Ray Positional Encoding for Multi-View Attention

Research·Apple Machine Learning Blog·Jul 20
A

Head of US AI safety agency resigns

Policy·CNBC +2·Jul 20
CRN

Alibaba Rallies After Unveiling New Qwen AI Model

Models·TradingView +2·Jul 20
TYG

Alibaba's Qwen Push Rattles China AI Rivals

Models·Yahoo Finance +1·Jul 20
YT

Xi offers four observations on AI development, governance

Opinion·CPPCC National Committee +1·Jul 20
CC

Moonshot unveils Kimi K3 and sparks debate over open LLMs

Other·Межа. Новини України.·Jul 20
М

Controlling Reasoning Effort in LLMs

Other·Hacker News·Jul 20
H

Beijing’s Campaign to Define the AI Age

Other·Homeland Security Today·Jul 20
H

American A.I. Giants Face Fresh Tests

Other·The New York Times·Jul 20
T

Airbus Takes Flight from AWS

Other·Hacker News·Jul 20
H

Trump’s latest AI czar has already resigned

Other·TechCrunch AI·Jul 20
T

AI Has a Language Problem

Policy·Center for European Policy Analysis (CEPA)·Jul 20
C

Sakana AI

Other·Sakana AI·Jul 20
S