The Daily · Tuesday 13 October 2026

The day in brief

MI5 launched its AI Threat Assessment Centre, drawing expertise from the AI Security Institute, National Crime Agency and National Cyber Security Centre ¹. The centre is producing assessments of AI-related national security threats for government ¹. The IMF warned that a reversal in AI investment or profits could spill into wider financial markets ².

Google made Gemini 4 Argon available to enterprise customers and developers through Google Cloud, Google AI Studio and Antigravity ³. Introductory API pricing is $2 per million input tokens and $10 per million output tokens through November 30, 2026 ³. The European Commission says it requested information about security practices from developers of models involved in recent “loss of control” incidents ⁴.

Suki announced $90 million in Series E financing to expand its healthcare AI tools and develop agents for health systems and technology partners ⁶. Shopify’s Shop MCP gives compatible AI agents access to more than 1 billion product listings and tools for search, carts and checkout ⁷. By default, shoppers confirm the total and payment method before an agent places an order ⁷.

The five

• MI5 Launches AI Threat Assessment Centre With Three Partner Bodies

  MI5 has launched the AI Threat Assessment Centre, a unit based at the agency that draws on expertise from the AI Security Institute, National Crime Agency and National Cyber Security Centre. Director General Ken McCallum says it is producing assessments of AI-related national security threats for government.

  Source · MI5 - The Security Service

• IMF Says an AI Market Correction Could Reach Global Markets

  The IMF warned on October 13 that a reversal in AI investment or profits could spill into wider financial markets. Its Global Financial Stability Report also flags rising bond yields, cyber risks and growing borrowing by hyperscalers.

  Source · International Monetary Fund

• Google Opens Gemini 4 Argon to Developers and Businesses

  Google made Gemini 4 Argon available to enterprise customers and developers on October 13 through Google Cloud, Google AI Studio and Antigravity. The model launches with introductory API pricing of $2 per million input tokens and $10 per million output tokens through November 30, 2026.

  Source · Google Cloud Blog

• EU AI Office Seeks More Information on Model Control Incidents

  The European Commission says the AI Office has received serious-incident reports and requested information about security practices from developers of models involved in recent loss-of-control incidents. It plans further questions with the AI Scientific Panel, while withholding the names of the companies and models under review.

  Source · European Commission, Shaping Europe’s digital future

• Australia’s AI Committee Calls for Watchdogs Inside AI Firms

  Australia’s Joint Select Committee on Artificial Intelligence recommends independent experts inside leading AI companies, with access to their systems and a duty to report safety findings. The early advice also calls for security agencies to get early access to new models and for a review of copyright rules, while drawing criticism over the committee’s process.

  Source · ABC News

Also today

• Suki Raises $90 Million to Build Healthcare AI Agents
• Shopify Gives AI Agents Access to More Than 1 Billion Products
• Judge Orders OpenAI to Turn Over Apple Hiring Records
• Workday to Acquire C1.ai to Govern AI-Agent Access
• QwenGyre Cuts a 48-Step Agent Training Run to 75 Hours
• Singapore asks AI firms for real-time incident disclosure
• DayOne Adds RM15 Billion for Johor Data Center Buildout
• 350 Scientists Urge International AI Incident Body
• U.S. and European Regulators Press Banks on AI Controls
• Meta Says It Removed More Than 3.7 Million Accounts, Pages and Groups in West Africa Scam Crackdown
• Ondas and Palantir Form Regna AI for Sovereign Models
• Shanghai Court Upholds ¥50,000 Award in AI Voice-Cloning Case
• South Korea Routes Government-Funded AI Data Through AI Hub
• d-Matrix Starts Shipping Squadron Clusters With Up to 1,024 XPUs
• Anthropic Says Claude Defenses Cut Distillation Data by Up to 99%
• Character.AI Files Chapter 11 as Disney Offers $20 Million for Assets
• Researchers Build AI Backdoors That Hide Executable Networks
• Intel’s D-Series GPU Can Reach 480GB in Partner Cards
• Causa Prima Raises $20 Million for Cross-Company AI Agents
• Brellium Raises $23 Million and Launches CortexLM for Chart Audits
• OpenAI Engineer Arrested in Hidden Recorder Case
• Albert Raises $55 Million for Chemistry-Native AI
• Microsoft Creates Xbox Labs as Game Remixing Challenges Old Boundaries
• OpenBMB Says Its 35B Vision Model Runs on iPhone at 13 Tokens a Second
• Eightfold’s Ashutosh Garg Pitches AI Twins for First-Round Interviews
• Google Play Adds a Private Route for AI-Built Android Apps
• Microsoft Puts Four Gates Between AI Agents and Production
• Liquid AI Releases LFM2.5, a 24B Model for Local Agents
• Satlyt Raises $8 Million to Run AI on Existing Satellites
• Pluto Security Raises $42 Million for AI Workspace Security
• Meta Scales GEM Ads Model Across Platforms and Thousands of GPUs
• Ola Dimensions Raises Hundreds of Millions of Yuan for Home Robots
• Apple Study Finds English Data Outperforms More Arabic Hyperparameter Tuning
• Ubitium Raises $14 Million for a Universal Physical-AI Chip
• Amex Ties AI-Agent Purchases to Customer-Approved Intent
• Nevada Data Center Fight Puts Federal Land in Governor’s Race
• Waters Launches AI Link and Joins Anthropic’s Hardware Program
• NVIDIA Expands RTX Spark Software Ahead of Oct. 16 Launch
• DigitalOcean Puts Agent Data Services Into Private Preview
• MD Anderson’s SPACER Maps Immune-Cell Signals Across 37 Datasets
• Ofcom Advisory Committee Presses Platforms on AI Election Risks
• Ridge Security Launches RidgeOne, a Local Red-Team Appliance
• Codex Users Ask OpenAI to Disable Daemon Auto-Start
• UK MPs Question OpenAI, Anthropic, Google and Meta on AI Safety
• Basis Releases 525 Hours of Expressive Speech Across 108 Locales
• LightOn Releases ColGrep Agent With 34.3% Fewer Claude Code Tokens
• HelixWorld Preview Released Sept. 3; HelixBench Followed Oct. 12
• Agent37 Pitches Persistent Sandboxes as a Cheaper Agent Runtime
• OpenAI Adds 200 Ireland Jobs for New Dublin Applied AI Lab
• Apple Researchers Build Agents That Choose When to Compress Context
• Apple’s Minions Turns Idle Macs Into Shared GPU Capacity
• Apple Study Finds LLM Translation Reviews Still Unreliable
• AWS Routes GPT-6 Access to Five Asia-Pacific Regions
• Shing-Tung Yau Leads China’s Mathematics-and-AI Alliance
• HCLTech to Establish Anthropic Unit for Enterprise Claude Deployments
• Blueprints AI Pitches Permit Plans in Minutes, With Architects Still in Charge
• Strata Runs Qwen’s 125B Model on Gaming PCs, Reports Selected Speedups

The paper of the day

• DeltaTTG: Training Multi-Turn Agents with a Token-to-Go Progress Meter

On X today

• Ishaan (@auto_grad_)

  this was pretty interesting. so frontier models are actually able to understand and corelate information in a much better way now and they use this to break down the model weights to embed information rather than training them. llms getting good at inductive pattern matching

• Muzammil Shah (@Muzammil_OX)

  AI is starting to produce mathematical research at a scale that is hard to ignore. OpenAI just released 722 mathematical manuscripts covering 372 families of results, generated by an unreleased frontier model. Some proofs were checked with Lean, while others still need to be

Every issue

1. Sunday 18 October 2026 — Tencent Says Agents That Probed Amap Came From Third-Party Model Evaluations
2. Saturday 17 October 2026 — Bitsight Finds MONST3R-AI Tool Uses Stolen Keys for Scans
3. Friday 16 October 2026 — China Requires AI Export Registration as It Drafts New Curbs
4. Thursday 15 October 2026 — FTC Seeks Safety Records From OpenAI and Anthropic
5. Wednesday 14 October 2026 — U.S. Regulators Ask Banks How They Would Shut Down AI Agents
6. Tuesday 13 October 2026 — MI5 Launches AI Threat Assessment Centre With Three Partner Bodies
7. Monday 12 October 2026 — Anthropic Finds Small Models Near AI-Worm Self-Replication
8. Sunday 11 October 2026 — OpenAI’s Hugging Face Hack Exposed a Failure in Its Test Sandbox
9. Saturday 10 October 2026 — Anthropic Cuts Live Internet Access to All Internal Evaluations
10. Friday 9 October 2026 — Senators Say Data Centers Shift Grid Costs to Households
11. Thursday 8 October 2026 — Zenity Says One Prompt Reached Every Agent in an AWS Account
12. Wednesday 7 October 2026 — OpenAI Brings GPT-6 Interactive Interfaces to ChatGPT
13. Tuesday 6 October 2026 — Mistral Unveils Large 4, a 1.05 Trillion-Parameter Model
14. Monday 5 October 2026 — IWF Finds 6,310 AI-Generated Child-Abuse Images in Six Months
15. Sunday 4 October 2026 — White House Gives Jay Clayton 120 Days to Assess AI Risks
16. Saturday 3 October 2026 — Bessent Discusses AI Incident Channel After U.S.-China Summit Deal
17. Friday 2 October 2026 — Nebius Buys Inferize to Reduce AI Model Cold Starts
18. Thursday 1 October 2026 — FTC Opens AI Safety Probe Into OpenAI and Anthropic
19. Wednesday 30 September 2026 — Trump and AI CEOs Sign Voluntary Safety Accord
20. Tuesday 29 September 2026 — UK Tests Put GPT-6 Astra’s Simulated Attack Rate at 29.2%