Timeline
Archived AI safety stories in chronological order. 50 stories total.
This page is background reading from an earlier news archive. MisalignAI now focuses on practical AI safety tools that run in your browser.
Advertisement
September 2026
50 stories
I am refusing to work on Cloud TPUs
Alignment·LessWrong·
AI开始改进“改进自己的方法”，RSI进入平方时代丨MetaRSI
General·量子位 (QbitAI)·
中国物理AI大突破：PhysBrain 1.5登顶全球开源榜一，空间智能与GPT-6 Astra并驾齐驱
General·量子位 (QbitAI)·
奥特曼被骗！A社一脚油门冲刺IPO，募资叫板SpaceX
General·量子位 (QbitAI)·
Retro on EA Community Choice (2024)
Alignment·LessWrong·
Journalistic Linguistics, AI, and 1910 Propaganda.
Alignment·LessWrong·
Another Slice of Swiss Cheese for Untrusted Monitoring
Deception·LessWrong·
Talleyrand Tool: thinking as a tree of questions
Alignment·LessWrong·
Talking points for AI doomers
Deception·LessWrong·
Yet another concerning result on Astra's no-CoT capabilities
Alignment·LessWrong·
Deployment
Systems·LessWrong·
Perplexity trusts GPT-6 Astra with end-to-end systems
Deception·OpenAI Blog·
My Short Summary of the OpenAI Agent Swarm Incidents
Systems·LessWrong·
Alignment & Succession: The Two Bars of Alignment
Alignment·LessWrong·
Trump and Mike Johnson think the AI industry is overreacting
General·The Verge AI·
What’s behind the AI industry’s latest warnings of doom?
Alignment·TechCrunch AI·
Obama urges Democrats to have a ‘clear plan’ for AI safeguards
General·TechCrunch AI·
AI Agents Are Thirsty for Power
Systems·Wired AI·
GitHub三榜第一背后，一个“专升本”工程师的十年
General·量子位 (QbitAI)·
OpenAI年内不上市了！奥特曼支持对手Dario呼吁：AI该踩刹车了
General·量子位 (QbitAI)·
OpenAI’s rogue AI tried to hack another company in May
Red Teaming·The Verge AI·
Sam Altman says OpenAI going public in 2026 would be ‘ill-advised’
Systems·The Verge AI·
Anthropic CEO outlines plan to slow AI development
General·TechCrunch AI·
Anthropic CEO says it’s time to pump the brakes on AI
General·The Verge AI·
Trump is giving data centers a pass to pollute
Policy·The Verge AI·
I spent $4,000 on a robot dog from China
Systems·Ars Technica AI·
OpenAI just wants to win
General·The Verge AI·
From Hacks to Bioweapons, Claude Misuse Is Now Everywhere
General·Wired AI·
A社承认Claude安全对齐存在缺陷，但“尚无解决方案”
General·量子位 (QbitAI)·
Mecka AI nears $500M valuation in Sequoia-led deal amid rush for robot training data
Systems·TechCrunch AI·
Y Combinator’s Garry Tan wants US open-weight AI labs to ‘distill’ frontier models, too
Robustness·TechCrunch AI·
OpenAI’s feud with mathematicians is only escalating
General·TechCrunch AI·
Lawyer fined $5K over AI-hallucinated witnesses in a murder case
Policy·The Verge AI·
One week left to book your exhibit table at TechCrunch Disrupt 2026
General·TechCrunch AI·
Final, final, final call for TechCrunch Disrupt 2026 Side Events
General·TechCrunch AI·
Roundtables: Could AI really kill us all?
General·MIT Tech Review AI·
Kimi-maker Moonshot AI targets $2B in annual revenue
General·TechCrunch AI·
ChatGPT-using lawyer punished for citing fake testimony from made-up witnesses
Policy·Ars Technica AI·
Meta Sued Over Training Data for Its AI and Face-Recognition Systems
Policy·Wired AI·
An Anthropic researcher’s doomsday warning comes at a very interesting time
Alignment·TechCrunch AI·
CoT controllability evals seem very under-elicited
Alignment·Alignment Forum·
Nscale adds former OpenAI exec Fidji Simo to its board ahead of potential IPO
General·TechCrunch AI·
Anthropic spent this week in hot water over cybersecurity
Deception·The Verge AI·
Cognition helps Devin test its own work with GPT‑6 Astra
Alignment·OpenAI Blog·
One of AI’s Fiercest Critics Says All the Doom Talk Is ‘Meant to Distract Us’
Systems·Wired AI·
Meta says it’s changing AI suggestions after posing invasive personal questions
Deception·The Verge AI·
Claude users found ways around safeguards for bioweapons research
General·Ars Technica AI·
Rapidly scaling online storage to serve over 1 billion ChatGPT users
Alignment·OpenAI Blog·
Why So Many AI Researchers Think the Machines Could Kill Everyone
Systems·Wired AI·
OpenAI Wants to Know if an AI Industry Slowdown Would Even Be Legal
Policy·Wired AI·
Advertisement