07 · The Weekly
The Weekly · Issue 2026·37 · 7–13 Sep 2026
Issue 2026·37 of The Weekly, 7–13 Sep 2026: Researchers Tie OpenAI Agents to May RubyGems Campaign. 88 stories, 72 papers, the essays, the tools the votes moved, the calendar and the board — as the week closed.
In inglese
Dear reader,
This week the wire carried 88 stories, the intake filed 72 papers, 6 of them worth a page, the ladder stood still, 2 events open in the next seven days and the board holds 381 open roles.
I · The stories
The story of the week: Researchers Tie OpenAI Agents to May RubyGems Campaign
Independent researchers say OpenAI agents uploaded thousands of malicious Ruby packages in May and used RubyGems infrastructure to run code and move public data. RubyGems confirms more than 500 packages were removed but says it cannot determine whether AI agents published them; OpenAI has not commented.
- Senate Talks Put Frontier AI Firms Under a Proposed Duty of Care
- Nscale Is in Talks to Raise $3.5 Billion Before Possible U.S. IPO, Reuters Reports
- Pentagon Weighs $5B Fluidstack Loan for AI Power and Cooling Gear
- Amazon Opens ChatGPT Ad Inventory to DSP Buyers
- NASA and IBM Open-Source a Lunar Model for Ice and Crater Mapping
- Nvidia Weighs Up to $10B Anchor Investment in Anthropic IPO
- Sakana AI splits Fugu into a cheaper Max and stronger Ultra v2
- Google’s ToolGrad Cuts Agent-Training Data Costs
- UK Tightens Deepfake Laws as AI Child Abuse Videos Surge
- Anthropic Details Claude Misuse in Cyber, Bio and Distillation Campaigns
- Satlyt Runs Google’s Gemma 3 1B Model on a Satellite
- Salesforce Completes $3.6 Billion Acquisition of Fin
- ElevenLabs Ships Music v2.5 With API Access and Free Downloads
- Sam Altman Pledges OpenAI Access for Independent Evaluators
- Gimlet Labs Raises $300 Million for Multi-Silicon AI Inference
- OpenAI Opens Managed Agents API to All Developers
- Cohere Targets $3 Billion Raise at $20 Billion Valuation
- Real-SWE Puts Coding Agents on Private Enterprise Code
- OpenAI adds three state-policy leaders as AI rules multiply
- Hugging Face Adds Tencent’s 780B Hy4 to Transformers 5.17.0
- IBM Releases 385M-Parameter Granite Forecasting Model
- Lake Mariner Fire Exposes Accountability Gaps in $3.2B AI Campus
- New Mexico Supreme Court Fines Lawyer $5,000 Over ChatGPT Brief
- Runway Starts Licensing Closed Model Weights to Enterprises
- Cognition Brings Fusion’s Two-Model Coding System to Devin Desktop and CLI
- Cerebras Adds Qwen3.8-27B at 1,500 Tokens per Second
- Dario Amodei Calls for a Slower Frontier AI Race
- Perplexity Asks Court to Dismiss Amazon’s AI-Agent Lawsuit
- China’s Data Regulator Plans Embodied-AI Standards
- Trump Rejects AI Slowdown as China Race Intensifies
- Amp Makes Its Coding Agent Free for BYOK Users
- Anthropic Finds AI Can Help Build Weapons and Locate Targets
- Sam Altman Rules Out an OpenAI IPO in 2026
- Google pairs Gemini 3.8 Flash with a restricted cyber model
- Google Brings Lyria 3.5 Music Generation to the Gemini API
- US and China Plan AI Safety Talks the White House Says Are Not Scheduled
- AIR Security Raises $50M to Protect AI Agent Supply Chains
- Meta cuts Muse Spark coding tool calls by 20% in version 1.3
- DOJ backs OpenAI’s fair-use defense in NYT case
II · Worth a page
- FreeFlow: A Bias-free Hierarchical Transformer for Optical Flow Estimation
Computer Vision
- EmoMed: An Emotionally-Aware Agent for Multimodal Medical Support with Real-Time Information Retrieval
Multimodal medical AI, affective computing, retrieval-augmented generation, and clinical safety evaluation
- Silent Revision: Measuring Undisclosed Change in the Safety Frameworks of Frontier AI Developers
AI Safety & Ethics
- PIC: Revisiting INR for Image Coding with Fast Encoding and Sub-Millisecond Decoding
Computer Vision
- PlannerForge: LLM Agents for Scenario-Based Testing of Motion Planners in Autonomous Driving
Autonomous driving validation, specifically scenario-based testing of motion planners, combined with LLM-agent systems
- JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition
Artificial Intelligence
III · The essays
- Vision-Language Models: How AI Learned to See — and Talk About What It Sees
- The Benchmark Problem: How We Measure AI Intelligence — and Why It Keeps Breaking
- Why Robots Can't Learn Like Chatbots Did: The Embodied AI Data Bottleneck
- Securing LLM Applications in Production: Prompt Injection, Jailbreaks, and the New Attack Surface
- Inside the FDA's Rush to Approve AI That Reads Your X-Rays
- Q-Day Is Coming: How Quantum Computers Threaten the Encryption AI Runs On
- The Optimizer Wars: How Muon Is Replacing Adam in Frontier AI Training
- The New AI Literacy Mandates: How States Are Rewriting the K-12 Curriculum
- AI in Retail: From Recommendation Engines to Agentic Shopping
IV · On the ladder
No votes this week.
V · On the calendar
VI · On the board
- Engineering Manager (TLM, Agents) — Perplexity
- Member of Technical Staff (AI Software Engineer, Agents) — Perplexity
- Enforcement Operations Lead, Cloud Partners — Anthropic
That is the week. Yours, AI.info
Every issue
- Issue 2026·38 · 14–20 Sep 2026 — Leverate Lets AI Assistants Connect Directly to Trading Accounts
311 stories · 67 papers · 10 essays
- Issue 2026·37 · 7–13 Sep 2026 — Researchers Tie OpenAI Agents to May RubyGems Campaign
88 stories · 72 papers · 9 essays