Harvard Kennedy School authors constructed four scenarios that combine assumptions about productivity growth, job displacement, and how income gains are distributed between capital and labour and across the income spectrum.
The South Korean government allocated 10 trillion won for AI-related budgets in 2026 and, on 28 August 2026, selected three consortia led by SK Telecom, KT and Kakao to operate “AI for All” and receive a total of 512 NVIDIA B200 processors during 2026.
Hugging Face reports that China’s monthly ceiling ran between 754B and 2.78 trillion parameters and U.S. models stayed under 130B in five of seven months, with the exceptions of Inkling from Thinking Machines Lab and NVIDIA’s Nemotron 3 Ultra at 561B in May and June, and that Qwen-based models account for 151,448 derivatives on the Hub, 2.6 times Meta’s total footprint and 4.7 times the Llama repositories.
Anthropic reports evidence that high-tenure users have developed habits and strategies to better harness Claude’s capabilities, with people using it for 6 months or more having 10% fewer personal conversations and a 6% higher education level reflected in their inputs.
Authors of “Who Bears the Cost of Honesty? A FAccT Workshop Synthesis and Research Agenda for Equitable AI Disclosure” report that workshop participants surfaced concerns that the use of accessibility-related AI could be held against workers in performance evaluations, and explored how context-specific disclosures may mitigate risks of stigma and misinterpretation.
Today’s lead: Half these jurisdictions have no AI law of their own.