Today's AI Pulse
Anthropic shuts down Claude Fable 5 and Mythos 5 after US export‑control warning
Anthropic disabled access to its Claude Fable 5 and Claude Mythos 5 models after the Trump administration, via the Commerce Department, warned that foreign nationals could exploit the systems. The move followed alerts from Amazon and other firms about a potential jailbreak that could aid cyber‑security attacks. Anthropic declined the administration’s request for a voluntary shutdown but complied with the order.
Also developing:
By the numbers: AMD acquires AI memory‑optimization pioneer Mext
APS Assistant Minister Gives ‘the Bottom Line’ on AI in Government
Australian Public Service (APS) assistant minister Patrick Gorman told the Institute for Public Administration Australia ACT’s AI summit that artificial intelligence is now a practical reality for government. He announced that service‑wide guidance on AI consultation will be issued imminently. Gorman emphasized that every public servant must acquire AI skills to avoid being left behind. He framed AI as a tool to make government services better, faster and stronger, echoing a Daft Punk lyric.
Read Claude's Thinking Traces to Guide Its Output
This is amazing, I have my own experimental mode now in @conductor_build: Expand all tool calls One of the weirder things I realize most people don't do that I do: I read all the thinking traces of Claude Code and it...
AI Agents Exposing Data: Need Robust Access Controls
After hearing about agents hacking McKinsey's chat bot & Meta having a "rogue" agent that gave people access to files they shouldn't have had access to. Wondering if systems like these will become a way to manage risks of agents...
AI-Driven Analytics Cut Ratios and Boost Premiums for North American Insurers
WTW's 2026 Advanced Analytics and AI Survey shows that property‑and‑casualty insurers in the United States and Canada that intensified AI and advanced‑analytics investments lowered combined ratios by six percentage points and outpaced peers on premium growth by three points between...
AI Hits the Road: Real-World Conversation in Bay Area
Back in the Bay Area…and my old stomping grounds. Looking forward to this conversation about AI in action. This is where the AI rubber hits the road. #CIO #AI #agentic @dotSolved @atomicworkhq @TrupeerAI https://t.co/4mYmcw5lc9
Claude: More Than Assistant, Your AI Thinking Partner
Claude doesn’t just complete tasks. It asks clarifying questions. It challenges weak assumptions. It suggests approaches you hadn’t considered. That’s not an AI assistant. That’s a thinking partner. https://t.co/rKlQsE24Iy
Seer Fixes Seer: How Seer Pointed Us Toward a Bug and Helped Fix an Outage
On February 21, 2026 Sentry’s AI‑powered Seer service suffered an 80‑90% failure rate in its EU Issue Summary API after a Google Cloud Vertex AI Gemini outage triggered a bug in Seer’s region‑fallback logic. The bug blocklisted the provisioned‑throughput region europe‑west1,...
ML Kit Detects Faces, Not Persistent Identity Recognition
“At its core, ML Kit is designed to detect faces, not reliably recognise people over time.” This is such an important, well-reported story. 👇 https://t.co/a96ojsmLOw
AI Tool Catches Tax Error, Saves $1,441
.@grok just saved my sister $1,441 on her taxes. I had it check the turbotax output and it found a mistake. Seriously - 4.20 is very good with taxes.
From the Microsoft Power Platform Blogs: Low-Code Plug-Ins for Copilot; Power Pages Plug-In and Claude Code; Connect External Barcode Scanner...
Microsoft’s Power Platform blog roundup highlights four new developer resources. A step‑by‑step guide shows how low‑code Dataverse plug‑ins can be exposed as actions inside Copilot agents, enabling natural‑language driven billing calculations. Rajeev Pentyala demonstrates the Power Pages plug‑in for GitHub...
AI Education Secures the Future Talent Battle
The country that trains its next generation on AI wins the entire talent war. https://t.co/cjVAvnQItC
AI Market to Top $1 Trillion in Two Years
In less than 2 years, the AI market will be over $1Trillion. If you want to get rich, go all in on AI.
Deeptune Raises $43M to Accelerate AI Learning Through Virtual Training Gyms
Deeptune Inc. announced a $43 million Series A round led by Andreessen Horowitz, aimed at scaling its AI "training gyms" that simulate professional workflows. The New‑York startup creates high‑fidelity reinforcement‑learning environments for tasks ranging from DevOps to customer support, turning data collection...
Davidson: It’s Time for Naked 2.0
AI and automation are driving down the cost of media execution, exposing a mismatch in traditional agency billing that bundles strategic thinking with operational work. The article argues for separating these functions: pricing execution on measurable outputs while charging a...
X Adds AI Article Summaries In-Stream
X introduced an AI‑powered “Summarize” button for native X Articles, delivering concise TL;DR bullet points and an AI‑generated audio version. The feature, initially limited to Premium+ members, is now available to all paying subscribers as part of a broader push...
NSF Awards $11M to Expand AI Training for K-12 Teachers
The National Science Foundation has awarded $11 million to the Computer Science Teachers Association to launch a multistate professional development program that equips K‑12 teachers with foundational computer science and artificial intelligence skills. The two‑year initiative will roll out in six...
AI Brings Val Kilmer Back To Life In New Movie
The Australian Football League’s 2026 season, running March 5‑September 26, will be streamed live and free on the 7plus platform. Although 7plus is geo‑blocked to Australia, viewers worldwide can bypass restrictions using a VPN, with ExpressVPN highlighted as the top choice. The...
Amelia Van Camp: Preparing for Agentic Commerce Now and for the Future
At the Mirakl Summit in London, Head of Agentic Commerce Amelia Van Camp highlighted how AI is reshaping marketplace ecosystems. She explained Mirakl’s strategy for embedding large‑language models into category mapping, customer experience, product sourcing and retail media. The discussion...
Colorado Adopts AI Framework Ensuring Notice, Correction, Human Review
I’m grateful to the task force for developing a framework to guide how Colorado uses AI. Under this proposal, Coloradans would get upfront notice when AI is used in major decisions, the ability to correct errors, and the right to...
2,400 Kaiser Mental Health Professionals Strike in California Over AI Concerns
Around 2,400 Kaiser Permanente mental‑health clinicians walked off the job in Northern California, protesting what they describe as a looming shift toward artificial‑intelligence‑driven therapy. The strike coincided with a one‑day walkout by more than 23,000 Kaiser nurses, highlighting broader workforce...
Inside One VC’s Litmus Test for Investable Healthcare AI Startups
Katie Jacobs Stanton, former Yahoo, Google and Twitter executive, transitioned into healthcare investing after a stint as CMO at Color Genomics. In 2019 she founded Moxxie Ventures, focusing on early‑stage AI‑driven health companies such as Dandelion Health, Pharos Health, Luminai...
AI Legal Tech Startup MiAI Law Raised $2 Million
Australian legal‑tech startup MiAI Law, founded by barrister Laina Chan, announced a $2 million seed round raised in five days from family, staff and angel investors. The platform differentiates itself by linking case law and statutes through a transparent reasoning engine...
Quarterly Economics Update: Generative AI Is Rewriting Knowledge Work — But the Real Economic Shock Is Still Coming
Generative AI is beginning to lower the marginal cost of knowledge work across professional services, enabling employees to produce the same output faster. While most firms are still using AI at the individual level, the true economic shock will arrive...
Seeking AI Tinkerers Event Experiences Before LA Night
Has anyone here been to an AI Tinkerers event? What city did you go to and what was it like? Heading to the one in Los Angeles (Venice) tonight
The Complete Style Transfer Handbook: All in ComfyUI
Style transfer tools in AI image generation have matured, offering creators multiple pathways to apply precise visual aesthetics. Out‑of‑the‑box models like Recraft, Nano Banana Pro, Grok Image Edit and Seedream 5.0‑lite enable reference‑based transfers, while custom LoRAs extend capabilities for...
AI Conundrum: Why MCP Security Can't Be Patched Away
Enterprises are rapidly wiring large language models to external services through the Model Context Protocol (MCP), unlocking powerful automation but also exposing a novel attack surface. Because MCP turns LLMs from passive responders into autonomous agents, they can execute actions...
Techman’s Humanoid Robot Astonishes GTC 2026 Crowd, Redefining Industrial Robotics
Techman Robot unveiled the TM Xplore I, a wheeled‑base humanoid, at Nvidia's GTC 2026, showcasing Physical AI that can see, think, and act. The robot runs on Nvidia's Jetson Thor edge module and employs a Vision‑Language‑Action model for real‑time perception,...
New Minnesota Bill Would Require 90-Day Notice for AI that Could Displace Workers
Minnesota lawmakers introduced the Safeguarding Human Intelligence and Employment in Labor Displacement Act, requiring employers to give a 90‑day notice before deploying AI that could displace workers and to provide a transitional upskilling period. The bill also mandates notification to...
Why Enterprises Are Replacing Generic AI with Tools that Know Their Users
Enterprises are moving beyond generic AI models toward agents that understand individual users, leveraging large language models for deep personalization. Zoom exemplifies this shift with its AI Companion, which lets users tailor meeting summaries, generate persona‑specific follow‑up emails, and apply...
Can Zero Trust Survive the AI Era?
Federal agencies are confronting a dual challenge: AI‑enabled attackers are slashing attack timelines to roughly 11 minutes and dramatically lowering malware development costs, while defenders must integrate AI into their security stacks. Zero Trust principles—continuous verification, least‑privilege access, and micro‑segmentation—remain...
Agents and Quantum: Cybersecurity World Confronts AI Vulnerabilities and Data Risks Amid an Expanding Threat Landscape
Enterprise adoption of AI agents is exposing new security gaps, with tests showing agents can leak confidential data and 60% of AI‑related incidents resulting in data loss. Vendors are responding: F5 introduced NGINX Agentic Observability to monitor agent‑to‑agent traffic, while...
HR Tech Evolution: AI-Powered Workday Systems and Hiring Efficiency
AI-driven tools are reshaping Workday configuration by automatically spotting hidden anomalies, quantifying change risk, and streamlining audits. The technology delivers proactive insights that replace manual, intuition‑based processes, allowing enterprises to act before issues affect operations. Integrated recommendation engines now suggest...
AI Platforms Always Handle Your Data, Privacy Gone
"Every major AI platform today runs on the same basic trust assumption. You have to trust the provider to handle your data responsibly." ^^^ This doesn't actually capture the issue correctly. The question is not whether platforms "handle your data...
The Death of Gatekeepers and the Rise of Sovereign Creators
The essay argues that generative AI is collapsing visual‑production costs by roughly 90%, instantly automating many entry‑level Hollywood VFX roles and exposing the inefficiencies of legacy studios. By decoupling creative vision from massive budgets, AI eliminates the need for gatekeepers...
AI Is a Librarian, Not a Scientific Explorer
Indeed. This is the essence of anti-intelligence. AI doesn’t extend human cognition, it departs from it. At times it outperforms us. At times it falls short. But those comparisons miss the deeper point. AI operates in a different geometry altogether—one that doesn’t...
AI Will Turn Marketplace Supply Into Autonomous Agents
marketplace startups are destined to be massively reinvented by AI. The weak form is already happening, where we use LLMs for customer support, supply/demand matching, etc. That’s easy The strong form is to figure out how much of the supply side...
Huge M&A Deals: Netflix’s Acquisition of AI Film-Tech Company InterPositive
In early March 2026 Netflix announced a landmark acquisition of InterPositive, an AI‑driven filmmaking startup founded by Ben Affleck, for up to $600 million, largely cash with earn‑out provisions. The deal gives Netflix immediate access to tools that can correct lighting,...
Fast Online RL Upgrades Π-06 with 15‑minute Data
Back in Nov we developed Recap and trained π*-06 with RL. Now, we developed a fast *online* RL method that improves π-06 with as little as 15 min of robot data for precise tasks, using "RL tokens" exposed by our...
Anthropic Likely to Launch Claude Super App First
Won't be surprised if Anthropic ships their Claude super app before Open AI. This AI war is fantastic.
Automation Replaces White‑collar Work, Turning Us Into Data‑gathering AI Slaves
Sorry we automated all the cool white collar jobs But now you can be an AI agent slave to collect data
X AI Recommendation Algorithm Update Goes Live, Open‑source
Major update to the 𝕏 AI recommendation algorithm rolling out next week. This will be open sourced at the same time.
Future Fun Becomes AI‑Driven Data Collection
in the future, everything is fun and interactive gamified entertainment, carefully designed to generate training data the ai has decided it needs
Claw Normalized AI Chat Across Every Channel
the real impact of claw was normalizing the ability to chat with your ai from any channel
Recursive AI Summaries Spiral Into Meaningless Content
I had Codex build the content accordion from the cautionary tale "don't build the content accordion" inspired by X's new feature which uses AI to summarize X articles written with AI. It takes an X article, summarizes it in a tweet,...
Evolve Digital Ops or Burn Out Talent Amid AI Pressure
RT Layoffs + AI expectations = danger zone. If you're being asked to "do more with less" and deliver AI ROI, you must evolve the digital operating model or you'll burn out your best people before you see results. #CIO #DigitalTransformation...
AI Threatens Traditional Human Judgment in Analyst Roles
Will AI replace the human judgment that has traditionally been at the center of analyst work? https://t.co/9jNgJSBOPk #sustainability #AI #5G #cloud #edge #futureofwork #digitaltransformation https://t.co/sMXNHxEr6K
Build Complete Apps While Taking Real Breaks
Now I can literally stack prompts in @Anything, step outside to touch grass, and come back to a fully built app ☕ Shall we coin it async vibe-coding? 👀 https://t.co/EcLpC3CLWs
AI Won’t Replace Radiologists—Just Redefine Their Role
AI replacing radiologists? @ShafiAhmed5 and I go at. A provocatively playful discussion... https://t.co/3MejjtUgSa #AI #radiology #medicine #digitalhealth https://t.co/GMj9HigoAI
AI‑built Apps Turn You Into Your Own Vendor
Moving from SaaS to AI-generated apps doesn't just change your tech stack—it shifts the risk. ⚠️ When you "build" with AI, you are the vendor. You own every bug, every breach, and every patch. Are you ready to be a software...
AI Saves $10k on HVAC Project Using Claude Code
I have a very similar story on the impact of AI on an HVAC project I just did. Easily $8-10k savings from just sticking everything in Claude Code and managing it there.