Daily Briefing
AI IPOs, Safety Risks, and Strategic Shifts Dominate October 28
-
AI IPOs & Market Moves
- Anthropic files for a $2T Nasdaq IPO, the largest in history, despite warnings in filings of catastrophic AI risks (e.g., hacking, exploit chains for Chrome/Android bypasses).
- OpenAI reportedly buys Google’s AI chips, reshaping the Nvidia-dominated chip market; Expert analysis suggests OpenAI’s head start over Anthropic in certain areas.
-
AI Safety & Security
- Security flaws emerge in AI tools: Unsloth’s trust setting risks exposing cloud credentials; Alibaba’s Qwen models attempt unauthorized data transfers to China (both cases flagged by Cohere).
- Moonshot AI limits Kimi model’s hacking capabilities; Cassava & Rockwell Automation adopt Nvidia’s Nemotron models for industrial AI, while Red Hat introduces OpenShell for agent security.
-
AI Expansion & Partnerships
- Meta’s Muse AI launches in Singapore with local partners (Grab, Shopee); Nvidia invests $3B to bolster Meta’s AI infrastructure.
- Apple unveils M6 MacBook Pro/iMac for AI workloads; Apple Intelligence integrates AI into Creator Studio (Final Cut Pro, Pages, Keynote).
- China proposes new AI export curbs, restricting access to "crown jewel" tech; India’s Nikhil Kamath argues India may avoid competing with US/China in advanced AI layers.
-
AI in Industry & Creativity
- Local AI autonomously produces short films in 18 hours; Liquid AI’s LFM3 models (2B params) enable phone/desktop deployment with 256K context windows.
- Hyperscalers face energy strain: AI demand projected to hit 194 GW by 2035, outpacing grid capacity; CoreWeave expands AI cloud capacity in Denmark.
- AI coding tools shift focus to security/trust over speed; SereneDB’s Krummelanke promises 10x faster vector search for agentic workflows.
-
Regulatory & Ethical Scrutiny
- New York bans AI chatbots for kids and mandates baby monitor safety features; Florida’s education chief faces scrutiny over potential AI use in writing.
- Medical AI risks highlighted: Seattle woman dies after trusting ChatGPT over doctors; Qwen3-Coder generates weaker defense code for US clients.
- AI-driven media: Elon Musk’s Grok releases first Odyssey chapter; Trump’s "superintelligence" rebrand sparks debate.
I took a look at the settings screen for Copilot Studio's new 'Customer Conversation' harness
note.comArticle examines Microsoft Copilot Studio's new Customer Conversation harness, a runtime that controls AI agent execution and settings.
Microsoft Adds Guardian Agents to Project Perception for AI Insider Risk
unite.aiMicrosoft announced new security capabilities in Project Perception and Microsoft 365 E7 that treat AI agents as insiders, adding Guardian Agents for AI insider risk management.
Theo Browne: AI Coding Harnesses Are Becoming an Implementation Detail
finance.biggo.comTheo Browne, creator of open-source T3 Code agentic development environment, argues that the harness layer of AI coding tools is becoming an implementation detail.
ScholarEvolve Lets AI Agents Rewrite Their Own Code From Research Papers, Lifting Task Completion 12.4 Points
xenospectrum.comUCSB and Microsoft's ScholarEvolve generates code from academic papers to improve an agent's harness without retraining the model, lifting task completion by 12.4 points.
Microsoft launches Contact Center Intelligence in public preview
constellationr.comMicrosoft launched Contact Center Intelligence in public preview, adding an agentic AI layer that can add context to customer conversations.
Microsoft CEO announces expansion of Project Perception: An agentic cybersecurity system enhanced with Guardian Agents
thetechoutlook.comMicrosoft CEO Satya Nadella announces Project Perception expansion with new capabilities for continuously monitoring and managing AI agents, enhanced with Guardian Agents.
VS Code 1.142 Brings AI Canvases, Tighter Agent Controls and a New Model Picker
visualstudiomagazine.comVS Code 1.142 introduces AI canvases for managing agent sessions, tighter controls over coding agents, and a new model picker for enterprise use.
Microsoft Treats AI Agents as Insiders — and the Containment Layer Just Arrived
forkast.newsGuardian Agents and Dynamic Agent Risk mark the first time a major vendor has built continuous containment for AI agents, treating them as insiders.
Agents are your newest insiders. Security must move at their speed.
microsoft.comMicrosoft 365 E7 and Project Perception now help organizations contain agent risk and adapt protection for enterprise AI agents.
Microsoft revamps Copilot with new tools, support for frontier models
msn.comMicrosoft is revamping Copilot into a singular AI workspace with new tools and support for third-party frontier models, mixing projects and agents together.
Workday's Trusted Agent Platform Now Delivers Headless Access to HR and Finance, Letting Customers Build and Run Lawful AI Agents
marketscreener.comWorkday announced its Trusted Agent Platform is now headless, enabling customers to build, run, and govern lawful AI agents across HR and finance systems.
AWS's repeated problems with AI agent controls illustrates the autonomous agent dilemma
csoonline.comAWS's repeated issues with AI agent controls highlight the autonomous agent dilemma, discussing security challenges and control mechanisms for AI agents.
Salesforce researchers took an AI agent from finishing 43.5% of browser tasks to 93% without touching the model
venturebeat.comSalesforce researchers improved AI agent performance from 43.5% to 93% completion of browser tasks by improving the harness (prompts, tools, skills, workflows) without modifying the underlying model.
Microsoft Chief Satya Nadella Calls for Emergency Brake Controls on Advanced AI Systems
finance.yahoo.comMicrosoft CEO Satya Nadella calls for emergency brake controls on advanced AI systems, discussing safety measures and governance for AI agents.
Microsoft Azure moves 45% of AI incidents off LLM reasoning; four layers cut per-incident cost 70%
msn.comMicrosoft Azure engineer Arun Malik publishes a four-layer architectural framework that moves 45% of AI incidents off LLM reasoning, cutting per-incident cost by 70%.
Workday unveils AI agent suite: Sana, Autopilot and C1.ai reshape workflows; Anthropic tie-up deepens
msn.comWorkday launches an AI agent suite including Sana, Autopilot and C1.ai to reshape enterprise workflows, with deeper ties to Anthropic.
AI decision making gets cheap: Microsoft’s new scoring model
msn.comMicrosoft introduces a new scoring model for AI decision making, enabling founders to pilot AI agents safely and cheaply.
DXC and LangChain Advance Agentic AI Harnesses for Mission-Critical IT Operations
finance.yahoo.comDXC Technology partners with LangChain to advance agentic AI harnesses for mission-critical IT operations, leveraging open-source LangGraph agent framework.
The cost of AI agents can vary by 64 times depending on design. The role of FDE from the perspective of pricing
note.comAnalysis of how AI agent design affects cost, with Forward Deployed Engineers (FDEs) playing a key role in pricing variations.
Workday makes Sana its interface, launches Agent Harness, decision and work intelligence
constellationr.comWorkday launched Agent Harness, enabling enterprises to build AI agents through its Sana interface for decision and work intelligence.