A LessWrong post by josefely argues the US and China can only durably slow AI through a treaty governing compute, because a unilateral pause would let China erase a roughly 10-point Epoch Capabilities Index lead. Verification would depend on satellite monitoring, on-site inspections and tamper-resistant chips that meter and license compute, the post says.
Sam Altman told OpenAI staff last week that the company has a year to pick one of three AI safety endgames, The Information reports. He described an IAEA-style global watchdog as his preferred option and called a single lab racing ahead to "melt everyone else's GPUs" terrible. He called the choice "pacing 2.0."
China's Ministry of Industry and Information Technology approves Alibaba's Qwen and Tongyi Wanxiang models for iPhones, alongside Apple Intelligence. Apple executive chairman Tim Cook met Vice Premier He Lifeng in Beijing. iPhone Duo pre-orders open October 16 across 70 nations, including China, shipping October 23.
Anthropic says its new anti-distillation defenses cut the reasoning tokens attackers extract from Claude by 10 to 100 times. The company's two-stage classifier, run on Claude Sonnet 5.5 and Opus 5.5, has a false positive rate under 0.1 percent. Anthropic also charges for requests blocked before Claude responds.
Microsoft CEO Satya Nadella says advanced AI systems need containment, independent controls and an emergency brake that lets authorized people pause or shut a model mid-task. Writing on X, Nadella says frontier models should be treated like insider risks. President Donald Trump opposes slowing US AI development and launched an AI Force led by Jay Clayton.
A 26-researcher working group that includes UC Berkeley professors Barry Eichengreen and Stuart Russell says an internationally verified pause on frontier AI training of at least 10 years is feasible. Their 200-page report proposes halting production of AI training chips and replacing them with inference-only chips that cannot be used to train new models. The researchers say success depends on world leaders' cooperation.
Manus raises more than $500 million in its first funding round since Beijing blocked Meta's $2 billion acquisition of the AI agent startup. Parent company Butterfly Effect says Boyu Capital and IDG Capital led the round, with Tencent, HSG and ZhenFund following on. Analysts say Manus still must prove profitability and regulatory alignment.
Anthropic updates its usage policy to ban using Claude for deceptive campaigns, weapons development, and election disruption, and requires a human in the loop for high-stakes uses. OpenAI says it shut down two influence operations, from Russia and Iran, that used its models for false-front campaigns. The Iranian operation, OpenAI says, created seven fake journalist personas.
The US-China AI gap narrows to a record-low 3%, according to Bloomberg Intelligence analyst Robert Lea, after DeepSeek's September release of V4.1 Flash. The model scores 77.3 on LiveBench agentic coding, ahead of Anthropic's Claude Fable 5.1 Max Effort at 66.1. Code sent to DeepSeek's API remains subject to disclosure under China's National Intelligence Law.
Mistral AI launches its 1-trillion-parameter multimodal Mistral Large 4, initially restricting access to a public guardrail endpoint before planning to release the model’s weights in three weeks following safety testing.