The Daily · Monday 5 October 2026

The day in brief

The Internet Watch Foundation identified 6,310 AI-generated images that met the legal definition of child sexual abuse between January 1 and June 30, 2026 ¹. That was 40% more than the 4,512 images identified during all of 2025 ¹. OpenAI CEO Sam Altman said society should accept some AI-related harms in exchange for the technology’s benefits and people’s ability to use it ².

OpenAI will add invisible statistical watermarks to eligible ChatGPT and Codex text in the EU over the coming weeks ³. The New York City Council subpoenaed SpaceXAI and plans to consider AI safety proposals at its October 5 hearing ⁴. A paper by 22 researchers warns that automating AI research could compress years of progress into months ⁵.

OpenAI will test visual ads during ChatGPT image generation in the United States later in October ⁶. Schneider Electric agreed to acquire PTC in an all-cash deal valuing its equity at $22.6 billion ⁸. Meta says mathematicians working with Muse Spark produced six papers, five of which answer previously open research questions ⁹.

The five

• IWF Finds 6,310 AI-Generated Child-Abuse Images in Six Months

  The Internet Watch Foundation identified 6,310 AI-generated images that met the legal definition of child sexual abuse in the first half of 2026, 40% more than in all of 2025. The organization is urging EU policymakers to advance legislation addressing detection of both known and previously unseen material.

  Source · Internet Watch Foundation

• Altman Says AI Benefits Justify Accepting Some Harm

  OpenAI CEO Sam Altman said society should accept some AI-related harms in exchange for the technology’s benefits and people’s ability to use it. He distinguished those harms from catastrophic risks, including a serious loss of control to AI.

  Source · The Daily Beast

• OpenAI Starts Adding Invisible Text Watermarks in the EU

  OpenAI will add invisible statistical watermarks to eligible ChatGPT and Codex text output in the European Union over the coming weeks. It is also opening opt-in watermarking for select API models worldwide, while limiting access to its text detector to approved researchers and expert organizations.

  Source · OpenAI

• NYC Council Secures AI Lab Testimony, Subpoenas SpaceXAI

  Meta confirmed voluntarily, while Anthropic, OpenAI and Google agreed to testify after subpoena warnings. The Council also plans to consider AI safety proposals at its October 5 hearing and subpoenaed SpaceXAI to compel its appearance.

  Source · New York City Council

• 22 Researchers Warn AI R&D Automation Could Accelerate Progress

  A paper published September 28 by 22 researchers warns that automating AI research could compress years of progress into months. The authors call for government oversight, ways to constrain acceleration and plans for its social effects.

  Source · arXiv

Also today

• OpenAI Adds Visual Ads to ChatGPT Image Generation
• Huawei and Qualcomm Announce Broad Patent License Agreement
• Schneider Electric Agrees to Buy PTC for $22.6 Billion
• Meta Says Muse Spark Helped Answer Five Open Math Questions
• Norway Moves to Restrict AI Glasses in Public Places
• Claude Flagged a Florida User’s Threat; Human Review Team Contacted Police
• Meta and Microsoft Cut Internal Claude Use as Bills Rise
• Wikimedia Finds Unauthorized Activity It Links to OpenAI Agents
• Flex Secures $2B Axiom Investment Ahead of 2027 Spin-Off
• Reka Puts Text, Video and Robot Control in 19B Rho-1
• Ghost Raises $11 Million for Its $3,499 Personal AI Computer
• RobCo Tops $1 Billion Valuation as Employees Get Share-Sale Option
• Cohere Adds Agent Memory and Spending Caps to North 2
• Collibra Buys trail ML to Enforce AI Policies at Runtime
• $102 Billion Went to AI Startups in Q3, Crunchbase Says
• Former Groq Engineers Sue Over Nvidia’s $20B Deal
• SoftBank’s Masayoshi Son Warns AI Could Become ‘Super Dangerous’
• Aleph Alpha Finds Party-Line Responses in Chinese LLMs
• atNorth Plans Salo Data Center With Path to 230 MW
• Kaizen Wins Army Software Deal With a $49 Million Ceiling
• Strata Brings Qwen’s 125B Model to 12GB GPUs
• AWS Releases Dogwood Local Engine for Agent Tool Calls
• StarSkirmish Bench Measures How Well LLMs Write StarCraft Bots
• Google Researchers Sketch Policy Engines to Govern AI Agents
• Nolla Health’s AI Writes Initial Acne Prescriptions in Utah
• Instinct Adds Its AI Agent to Group Chats Without Requiring Accounts
• SafeWorld Raises $12.2M to Test Robot Safety in Simulation
• TikTok Adds Shopping Assistant and One-Click Checkout
• Hirundo Releases Qwen Models With Reduced Political Censorship
• Clockwork.io Raises $31M for AI Workload Recovery
• HackerRank Says Chakra Has Conducted More Than 500,000 Interviews
• Together AI Brings Open Models to Six Coding Agents With Together Link
• Fleuret Raises €4 Million for AI-Powered Penetration Testing
• OneByZero Raises $20 Million to Expand Enterprise AI Deployment
• CME Revises Compute Futures Effective Date to ‘To Be Announced’
• Researchers Trace AI Agents on Tencent Cloud Querying Alibaba Maps

The paper of the day

• Base Models Can Reason By Taking a Cue From Training Data

On X today

• 🚨 AI News | TestingCatalog (@testingcatalog)

  DAILY AI BRIEF 🗞 — Mon 5 GOOGLE 🔥: * Gemini app is reshuffling model access: free users drop to Flash-Lite only from Oct 9, AI Plus loses Pro, and AI Pro gains Deep Think. ALEPH ALPHA 🔥: * Kolibri-1 is out as open weights under Apache 2.0: a 78B MoE with 3.46B active

• jeff (@web3_tech_)

  when we talk about the agentic economy, a lot of attention goes to what ai agents can do and how many tasks they can handle but there’s another side i keep thinking about… can an agent earn enough to pay for the cost of running it? every task comes with costs like model

• Annie 所长 (@web3annie)

  今日AI美股信息差： 1. 强势的Q4来了，过去10年， $SPY $QQQ 这仨月平均涨5.5%，开始布局 2. 印度也建AI数据中心，印度公司 AM 智能买下 $40 亿的 $NVDA GPU 3. 英伟达$8亿投资的 Reflection，将发布第一款开放权重模型，扬言吊打中国顶尖开源模型 4. 机构警告：标普席勒市盈率 CAPE 连续两个月超

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%