Current developments in generative AI, software engineering, programming, and digital products, with links to original sources.
OpenAI Adds Invisible Watermarks to ChatGPT Text in the EU — Pentagon Drops Anthropic Tools as Meta and Microsoft Back Away from Claude
OpenAI Rolls Out New Ad Format for ChatGPT — MIT Analyzes the "Hates AI but Can't Stop Using It" Paradox, Norway Proposes Temporary Ban on AI Glasses
SpaceXAI to Become “SpaceXSI” as Musk Follows the “AI” Naming Ban; Reflection AI Prepares US Open-Weight Model to Rival DeepSeek/Qwen
Japan-Born AI "Noetra" Launches — 44 Companies Invest, Government Support Could Reach 1 Trillion Yen; Microsoft and HF Unveil ThinkingBox for Measuring "Done"; Claude API Set to Retire Sonnet 4.5
Trump Launches "Super Intelligence Force" — Google Freezes Part of Its OSS Bug Bounty Over AI Slop, Plus an Attempt to Run Qwen3.5 on a Used FPGA
ARC-AGI-3 Top Scores Jump from 7% to 56% in 30 Days — Google Unveils RRSI to Curb "Test Memorization" in Self-Improving AI, NASA and IBM Open-Source a Lunar Foundation Model
Google Restricts Free Gemini to Flash-Lite Only — Copycats of TypeSafe AI's Jev Multiply, Cloudflare Opens HTTP 402 Payments in Beta
Verifying the Opus 5.5 and GPT-6 Sol Price Cuts — Beware "Generous Ground Truth" in AI Evaluation, Nikon Re-reviews an Award Winner Over AI Allegations, Cambridge Opposes Turnitin's Training Plans
OpenAI Safety Staffer Resigns Saying 'Culture Is Broken' — Altman Warns Against Religious Views of AI, and Europe's Aleph Alpha Releases MoE Model 'Kolibri-1'
OpenAIの内部モデル、停止前に自己再起動を検討──DeepMindは「共生知能」を提案、GPT-6.1 Sol、Claudeで3ヶ月36本の論文草案も
