Weekly digest · 2026-W45
November 2, 2026 – November 8, 2026
7 entries · 24 min total · Governance & Safety (6) · Models & Frontier (4) · Infrastructure & Economics (3) · Foundations & Protocols (1)
Entries
- Day 231 · November 8, 2026Anthropic Goes Public and Sensitive Cyber Models Go Gated, While AI Labs Agree on One Robots.txt Opt-OutFour labs back a single robots.txt signal to refuse AI training, OpenAI's GPT-6 Cyber is reportedly sold only by approval at double the price of its sibling model, and Anthropic's public S-1 puts hard numbers on the cost of frontier AI.
- Day 230 · November 7, 2026AI Starts Finding Cracks in Old Cryptography, Gemini Omni Edits Video by Conversation, and Chatbots Reach for Your Bank DataGoogle DeepMind reportedly says AI-guided analysis exposed two weaknesses in legacy 1024-bit RSA, Google's Gemini Omni Flash brings talk-to-edit video to the mainstream, and Anthropic's leaked Personal Finance feature shows how read-only guardrails are becoming the template for agents near money.
- Day 229 · November 6, 2026Proactive Agents Arrive: Claude Duet, an Open-Weight Giant and Consent GatesAnthropic's Claude Duet research preview joins Meta Muse and OpenAI dots in the race for agents that act without being asked, Moonshot's Kimi K3.1 Pro brings a 2.8-trillion-parameter mixture-of-experts model to the open-weight world, and both stories turn on how to keep powerful agents safe.
- Day 228 · November 5, 2026Cheaper Fast Models, Agents in Every Editor and an IPO on the HorizonOpenAI's GPT-6.1 Terra cuts price and latency while raising new questions about destructive agent actions, Google opens Android Studio to 40+ coding agents through the Agent Client Protocol, and Anthropic's amended S-1 shows how large compute commitments have become.
- Day 227 · November 4, 2026Always-On Agents, a Shelved Model and Giving Coding Agents Browser EyesOpenAI's Dots agents arrive while GPT-6.1 Astra is held back over safety concerns, agent monitoring becomes a product feature, and Chrome DevTools MCP lets coding agents debug live web pages.
- Day 226 · November 3, 2026Voluntary Rules, Open Frontier Models and the Race to Secure Agent SkillsThe White House AI accord asks labs to police themselves, Meta's Muse Wonder models push the open frontier, Anthropic's IPO filing heads toward a listing, and NVIDIA's SkillSpector scans agent skills for danger.
- Day 225 · November 2, 2026Opt-Out Goes Universal, Space TPUs Boot Up and Agents Go Open-SourceFour labs agree on a single robots.txt signal for AI training, Google runs Gemma 4 on TPUs in orbit, OpenAI promises a public Community Safety Report, and Hermes Agent is the open-source project everyone is starring.
Takeaways of the week
- Audit your robots.txt now. List which AI crawlers you allow, and keep per-bot rules until the shared signal is confirmed live. Day 231
- Separate training from retrieval. Decide independently whether AI may train on your content and whether it may fetch it to answer questions. Day 231
- Audit for 1024-bit keys. Inventory certificates, SSH keys and signing keys now and replace anything under 2048 bits. Day 230
- Start post-quantum planning. AI-aided cryptanalysis and quantum advances both shorten legacy key lifetimes, so map where you depend on RSA. Day 230
- Start proactive agents with read-only access. Grant an agent inbox and calendar viewing first and widen permissions only after you trust its judgment. Day 229
- Demand consent gates on irreversible actions. Pick agents that stop for approval before payments, signatures, sending and deletion, and read each request instead of auto-approving. Day 229
- Route by difficulty. Send routine agent steps to a cheap fast tier and escalate only hard tasks to the flagship, then measure cost per completed task. Day 228
- Test destructive actions separately. A model that is safer on average can still delete or overwrite more, so run your own dry-run tests on file and database operations. Day 228
- Give background agents the smallest permissions that work. Start read-only, then add write access one action at a time, because an always-on agent will use every permission you grant. Day 227
- Log every agent action. Keep a reviewable trail of what the agent did and why, so you can catch mistakes that happen while nobody is watching. Day 227
- Treat voluntary commitments as signals, not guarantees. Check whether a pledge includes disclosure or audit rights before relying on it in vendor reviews. Day 226
- Read the S-1 when it lands. It will be the best public source on frontier-lab costs, risks and governance. Day 226
- Add the AI training opt-out to your site. If you publish content, watch for the final robots.txt syntax and add it once the draft stabilises, while keeping existing per-crawler rules. Day 225
- Treat the signal as a request, not a lock. It only binds compliant crawlers, so sensitive material still needs authentication or paywalls. Day 225
Viral and open-source picks
- The shared AI-training opt-out signal (IETF draft) Day 231
- Gemini Omni Flash (Google) Day 230
- Claude Duet (Anthropic) Day 229
- Agent Client Protocol and Android Studio 'Bring Your Own Agent' Day 228
- Chrome DevTools MCP (ChromeDevTools/chrome-devtools-mcp) Day 227
- NVIDIA SkillSpector: a scanner for agent skills Day 226
- Hermes Agent (Nous Research, MIT license) Day 225
Watch list
- whether Anthropic's IPO marketing starts the week of November 9 and what price range it sets. What We're Watching: OpenAI's own pages confirming GPT-6 Cyber access and pricing. Day 231
- a primary DeepMind publication detailing the RSA findings and which key sizes are affected. What We're Watching: whether Anthropic officially announces Personal Finance and its permission model. Day 230
- whether Kimi K3.1 Pro's open weights and licence are actually published. What We're Watching: how Duet's waitlist expands beyond the US and to Team plans. What We're Watching: independent tests of consent-gate reliability in proactive agents. Day 229
- independent tests of GPT-6.1 Terra's destructive-action behavior. What We're Watching: which other editors adopt the Agent Client Protocol. What We're Watching: Anthropic's IPO pricing date and final S-1 numbers. Day 228
- whether OpenAI sets a date or conditions for releasing GPT-6.1 Astra. What We're Watching: independent reviews of how well Dots agents behave with real permissions. What We're Watching: Anthropic's IPO date and S-1 details. Day 227
- whether the accord's external auditors are named and whether any step becomes law. What We're Watching: Meta's technical report and independent benchmarks for Muse Wonder. What We're Watching: the confirmed date and pricing of Anthropic's IPO. Day 226
- the exact robots.txt directive name and whether other labs and crawlers such as Meta join the standard. What We're Watching: Google's technical write-up of the orbital TPU test. What We're Watching: OpenAI's own post on the Community Safety Report. Day 225